VisicalcWelcome to the first episode of Retro Tech Bytes,  where we take a look at the history of  a piece of software or hardware that was  either historically significant or just  has an interesting story attached to it.Today we will be looking at the history of  the very first personal computer spreadsheet,  VisiCalc. And for that, we need to go all  the way back to fall of 1977 at the Harvard  Business School where a first year student has  just encountered an incredibly tedious task.Dan Bricklin had been a programmer for about four  years, both as an employee of Digital Equipment  Corporation, a major player in the minicomputer  computing world of the 1970s as well as for  himself as an independent software developer.  However by 1977, Bricklin felt that programming  was actually getting to be so easy that he would  be out of a job in the future if he didn’t do a  career switch[a]. As hard as it is to imagine  this now, in the 1970s higher level programming  languages like COBOL, FORTRAN, and Smalltalk  had brought a new accessibility to programming,  especially when compared to the programming  world of the 1950s and early 1960s. Given the  comparatively greatly reduced barriers to  entry, Bricklin felt that the world would  soon have a marked oversupply of programmers, and  so he decided to enroll in the Harvard Business  School and get his MBA in preparation  for a new career in the business world.As a fresh MBA student in fall of 1977, Bricklin’s  production professor introduced him to production  planning using a matrix that could easily span  multiple blackboards across multiple rooms. This  matrix of rows and columns was used to create  what-if scenarios, where changing the value in  one cell; for example employee wages; sent changes  rippling across all related cells. This was great  for modeling various business cases, and was  a fairly common business tool at the time, but  it was also a brutally tedious one as every change  in a cell had to be manually recalculated by hand.  True, there were fairly cheap calculators around  by the late 1970s but each calculation still had  to be manually punched in, worked through,  and handwritten over an erased value on the  blackboard. Since a single change in a cell could  ripple like a mathematical wave across dozens of  other cells, large financial models could require  hours or even days to recalculate. And of course,  any mistake by the human doing the calculation  or any missed cells that were linked to  other cells would affect all other related  calculations and be difficult to track down.Bricklin wasn’t impressed by the thought of  wasting large chunks of his professional life  on these boring, repetitive, error prone  calculations. In his book “Bricklin on  Technology” he cites a paper he wrote in late  1978 for an advertising class that gives some  insight into his thoughts at the time, stating  that after spending hours working on homework  calculations he would “invariably find that  one of [my] initial calculations was in error,  invalidating all of the numbers that  followed it.” As a professional programmer,  his thoughts naturally turned to, as he put it  “If only I had a magic piece of paper where I  could change a number at the beginning of a set  of calculations, and have all of the other numbers  automatically recompute themselves...If only I  had an electronic spreadsheet.” and he started  thinking about using one of the new personal  computers to do these calculations automatically.This was not necessarily an obvious  idea, in 1978 there were only about  two hundred thousand personal computers worldwide.  The world’s first significant personal computer,  the Altair 8800, had only come out three  years previously and the heavy hitters  of the Commodore Pet, TRS-80, and Apple  II had just come out the previous year  and were still gaining traction. The Apple II  in particular was lagging the other two systems,  having only sold about 20,000 units in 1978.  Mainframes were embedded in companies large  and small, but they were largely used  for things like inventory management,  payroll processing, and things of a  similar nature, frequently batch related.Financial software did exist, including packages  like Foresight and Business Planning Language,  but these were more time share focused, were not  really interactive, required considerable training  to use, and were solely used on mainframes  designed to service many simultaneous users.  According to one source “Spreadsheet-type  applications were readily available on  mainframes and time-sharing computers (usually in  the form of financial modeling packages), but the  lack of interactivity limited their usefulness.”  They were primarily tools requiring professional  support to use and they were frequently not real  time. Additionally the timesharing systems cost  hundreds or even thousands of dollars a month  in fees to access, further limiting their reach  and usefulness. The name “spreadsheet”  did exist, although it typically referred  to the physical matrix based financial models  that Bricklin saw at Harvard Business School.The lack of interactivity in systems such  as the pioneering LANPAR system from 1969  greatly limited their usefulness since there  was no easy way to sit in front of a computer  screen and start playing with various financial  scenarios in real-time. Additionally, given that  all existing spreadsheet type applications ran  on expensive mainframes they were not widely  available to an individual but were rather  more the domain of large businesses. However  the new personal computers were orders  of magnitude cheaper than these systems,  and Bricklin realized that the demands of an  electronic spreadsheet could be handled by even  their far slower processors. A spreadsheet after  all is not a terribly complex system, at least not  in its most basic form, it is an excellent example  of a tedious task that is easily computerized.What Bricklin was envisioning was an easy to use  spreadsheet that was customized for a single user,  running in real time on a comparatively  inexpensive personal computer. His  program could not only be used for running  a small business’s finances but also for an  individual who just wanted to keep track of  their monthly budget. Bricklin originally  imagined his electronic spreadsheet  as part of a heads up display where he  “could see the virtual image hanging in the air in  front of me. I could just move my mouse/keyboard  calculator around on the table, punch in  a few numbers, circle them to get the sum,  do some calculations and answer ‘10% will be  fine!’” Bricklin quickly refined this imaginative  vision down to something that could realistically  be achieved by a 1 MHz processor with 32k of RAM.Bricklin received encouragement from some of  his teachers, as both his production professor  and his accounting professor loved the concept,  although his finance professor did not agree,  telling Bricklin that there were already plenty  of sophisticated financial analysis programs  available for mainframes and that there was no  room for a little one that ran on a personal  computer. However this professor did suggest  that Bricklin talk to a former student of his,  Dan Fylstra, who had been exploring the  possibilities of selling software to the  personal computer market, and had cofounded a  small software publishing company called Personal  Software after graduating from Harvard Business  school with his own MBA in 1977. Personal Software  was a pioneer in the early personal computer  software market, and was doing quite well,  thanks to an early chess program called MicroChess  created by its other cofounder, Peter Jennings.Bricklin originally created a very rough  prototype of VisiCalc in early 1978,  written entirely in BASIC on Harvard’s timesharing  system. It primarily served to help him flesh out  the concept further. This is where some sources  tend to mix up Bricklin’s first prototype,  created in early 1978 on the time sharing system  in BASIC, and his second prototype, which he did  in Apple BASIC on a borrowed Apple II in the fall  of 1978. The first prototype on the timesharing  system helped him refine the idea and matrix  interface and how the cells would be addressed,  the second one on the Apple II helped him work  out things like how navigation would work.Bricklin had originally hoped to use a mouse  to control things, since he was aware of the  pioneering work done by Douglas Engelbert and  Xerox PARC, but this was unfortunately not  possible on a 1978 era Apple II, which wouldn’t  see a mouse until a prototype of a mouse and  graphical user interface was created by Bill  Budge, Burrell Smith,and Andy Hertzfield in 1981,  later released in 1982. Four years away from a  viable Apple II mouse, Bricklin’s next idea was to  use the Apple II’s game paddles to hop from cell  to cell, but he found them to be too sluggish and  imprecise for the fast navigation that he wanted,  so he eventually went with using the arrow keys.  One wrinkle was that the Apple II did not have  up and down arrow keys, only left and right,  forcing Bricklin to utilize the space bar to  switch between horizontal and vertical movement.Having worked out how he wanted  VisiCalc to look and work,  Bricklin then partnered with a former roommate  of his from MIT by the name of Bob Frankston,  who was an accomplished programmer in his own  right. The two of them formed a small company  called Software Arts with the goal of developing  a full version of VisiCalc for retail sale.  Frankston would be the one to actually  program the full version of VisiCalc,  using Bricklin’s prototype as a reference and with  constant refinement and feedback from Bricklin.  A veteran programmer, Frankston had been  programming since the 1960s and his handling  of the coding allowed Bricklin to focus on  the high level view of exactly how all the  pieces of the program should work, as well as  start writing some of the user documentation.When Bricklin spoke with Fylstra, he was very  interested in VisiCalc and quickly offered to  publish VisiCalc through Personal Software.  Fylstra was also responsible for the choice  of the Apple II as the initial target platform.  The Apple II was chosen purely because Fylstra had  both it and a TRS-80 on hand in fall of 1978 when  Bricklin first spoke to him and since he wasn’t  using the Apple II, that’s the one he loaned it  to Bricklin, who then used it to create his second  prototype. This Apple II was solely used for  developing the second prototype, it was not used  to develop the actual retail version. Bricklin  had no personal preference between the TRS-80,  the Commodore Pet, or the Apple II as he owned  none of them and thus the enormous sales benefit  that Apple gained from being the initial VisiCalc  target platform happened purely by chance.Once the second prototype had been used to  work out all the implementation details,  Frankston got to work developing the final retail  version that would be distributed by Fylstra and  Personal Software. He implemented VisiCalc  entirely in assembler, using a time-sharing  system called MULTICS that ran on a big mainframe  that could be accessed remotely for an hourly fee.  Since the fee for computer time was cheaper at  night, VisiCalc was primarily programmed during  nights, with Frankston remoting into the mainframe  from his house, programming during the night and  sleeping during the day. Frankston wasn’t  remoting in from a personal computer either,  he did his development from a DEC LA-120 terminal  with no screen, just an endless roll of paper.  Given that Fylstra had loaned Bricklin an  Apple II, I would assume that Frankston  used the emulator on a timeshare system due  to the enhanced programming tools and more  reliable backups the mainframe he was remoting  into offered. Programming the early personal  computers was still difficult in the late 1970s  as they lacked many of the helpful tools that  mainframes had and suffered from frequently  unreliable storage. Although some sources state  that Frankston was using an Apple II emulator, he  in fact appears to have specifically used a 6502  emulator. Since he was coding entirely in 6502  assembler anyhow this would make the most sense.There seems to be some confusion over how long  it took to program VisiCalc as Wikipedia says  it was programmed in two months at the end of  1978 while the seminal work on the development  of the personal computer, Fire In The Valley, says  the first prototype of VisiCalc wasn’t available  until spring of 1979 and Accidental Empires  states that it took “close to a year” to finish.  This is presumably due to the confusion  over just how many versions of VisiCalc  were created. To clarify, there were  three separate versions of VisiCalc,  starting from when Bricklin first created a rough  prototype in BASIC on the Harvard mainframe, then  another version done in Apple BASIC on the Apple  II borrowed from Fylstra, and finally the full  version done in 6502 assembler, with Frankston  primarily working on a mainframe that ran a 6502  emulator. Final development was then moved to a  minicomputer that Frankston and Bricklin bought  when they first rented office space. This last  one was the first to be given the name VisiCalc,  as up until that point it had been known as  Calcu-Ledger. Fylstra, in his capacity as the  man who would market the product, chose the name  VisiCalc, short for Visible Calculations, during  a meeting with Frankston at a restaurant called  Vic’s EGG on One in Cambridge, Massachusetts.There is also some confusion over when it  launched with Wikipedia stating June 4th of 1979  and sources such as Accidental Empires giving the  date of October of 1979. This is because although  VisiCalc was first announced and publicly shown at  the National Computer Conference in June of 1979,  it initially attracted little interest and  there was still additional development needed  on it. However, an influential analyst by  the name of Ben Rosen saw a demonstration  of pre-release version of VisiCalc and  wrote that it was “hard to imagine any  serious user of a personal computer not  owning-and frequently using-VisiCalc.”  He additionally stated that VisiCalc “seems to be  unique in the computer industry. Mainframe people  I’ve shown VisiCalc to claim there’s nothing  like it available on conventional machines.”  His prediction would soon be put to the test, but  first VisiCalc needed to complete development.Based on VisiCalc’s anticipated sales, Personal  Software advanced Bricklin and Frankston a  prepayment of 100,000 dollars, enabling  them to rent space in the basement of a  commercial building, buy the aforementioned  minicomputer to finish development on,  hire two employees and finish VisiCalc’s  development. They also started the process  of porting VisiCalc to the other two dominant  members of the nascent personal industry,  the Commodore PET and TRS-80. The PET was an easy  choice since it used the same 6502 processor that  the Apple II did, which could not be said  for the TRS-80, which used a Z80 processor.Although VisiCalc was technically  available for sale at this point,  only a literal handful of copies seem to have been  sold, with Bricklin remembering only five copies  being shipped sometime that summer. VisiCalc’s  first real, boxed release was version 1.37 which  shipped in October of 1979 with an initial  retail price of 100 dollars. As released,  VisiCalc only ran on an Apple II with an expanded  RAM of 32k, eight times the 4k of the original  1977 Apple II and double the standard RAM of the  newly released Apple II+. VisiCalc was also a  pioneer in its packaging, with Personal Software  shipping the boxed release out in a handsomely  designed brown and gold binder, with excellent  documentation that was professionally typeset  and illustrated. The vast majority of software  in 1979 was merely a floppy disk or cassette tape  inside a ziploc bag with xeroxed documentation  that was merely stapled together. Fylstra knew  the importance of presentation, and made sure  that VisiCalc presented itself professionally.Although slow to start, VisiCalc quickly became  the first “killer app” of the personal computer  era. Up until VisiCalc’s launch, the uses for a  personal computer were not necessarily evident  to most people who weren’t programmers. The  Altair 8800 that kicked off the personal  computer industry in 1975 originally came with  no software, no display, had to be programmed in  machine language by flipping toggle switches  on the front, and had no persistent storage.  This all was rapidly changing but by 1979 the most  useful thing for most people that a computer could  do was probably word processing, with the first  word processor, Electric Pencil hitting the market  in 1976, and the WordStar reaching the market in  1978. But this was still a tiny market overall.Now however, for about 2500 dollars in  1979 dollars, a business could purchase  a copy of VisiCalc and an Apple II to run it  on and “have a turnkey system for handling  the accounting and spreadsheet work they had  been doing longhand (at small companies) or on  mainframes with punch cards (at big companies)”  accessing real time capabilities that even big  businesses lacked. And this was a single  one time cost, not a monthly recurring one.VisiCalc was something that any MBA, financial  analyst, small business owner, or anybody who  dealt with numbers on a daily basis, could  immediately grasp and see its enormous utility.  People bought a computer solely to run VisiCalc  and sales of Apple IIs, previously lagging far  behind the PET 2001 and the TRS-80, its two main  competitors, skyrocketed. Prior to this Apple  had been selling only small numbers of Apple  IIs but with VisiCalc fueling enormous demand,  Apple’s sales exploded even after VisiCalc’s price  was raised from 100 dollars to 150 dollars and  then to 250 dollars by 1982. Although VisiCalc  was only exclusive to the Apple II for a year,  it gave Apple the jump it needed to become the  powerhouse it became in the 1980s. Additionally,  because VisiCalc required the upgraded 32k Apple  II over the standard 16k model, Apple made even  more money per sale as the upgraded Apple II  had a higher profit margin than the base model.VisiCalc eventually was ported to most major  computers of the era, not just the Commodore  PET and TRS-80 but also Atari’s 8 bit line of  computers, CP/M machines, and was available at  launch for the IBM PC in 1981. Personal Software  was put in the somewhat unique position of being  primarily a single software company, almost  entirely dependent on VisiCalc for revenue,  a program that as the publisher, it didn’t own the  rights to. A number of other programs were created  to play off of the VisiCalc name, such as a GUI  called Vis-On and some programs that extended  VisiCalc’s functionality such as VisiTrend and  VisiPlot. These last two programs were purchased  by Personal Software and were also unique in  that one of the two programmers who created them,  Mitch Kapor, would later leave Personal Software  to form his own company called Lotus, whose  flagship product, Lotus 1-2-3, would play a key  role in VisiCalc’s future. More on that in a bit.1979 started with 500 copies of VisiCalc shipping  every month but that number quickly expanded to  12,000 copies a month by 1981, and then to  30,000 copies a month, reaching a total of  a million copies sold by 1985 when VisiCalc  was discontinued. However by this time it  had been essentially moribund since 1983, when  Lotus 1-2-3 came out for the IBM PC. VisiCalc  simply did not keep up with the rapid pace  of software development that the computing  industry of the 1980s required and although it  had been available for the new IBM PC in 1981,  no special effort had been made to take  advantage of the PC’s feature set. The  IBM PC 5150 could have up to 640k of RAM, but the  version of VisiCalc released for it was merely a  port of the TRS-80 port of the original Apple II  version with essentially no special development.There is some conflicting information here on how  much RAM the original IBM PC version supported as  Accidental Empires says that it could use no more  than 64k of RAM, and had the exact same feature  set as the old Apple II version while Bricklin  himself says that it could in fact make use of a  full 512k that an upgraded IBM PC could come with,  quite a lot of memory in those days, although less  than the 640k that a fully upgraded IBM PC could  possess. I am going to go with Bricklin’s account  here, which does not disagree that the PC version  of VisiCalc was feature-wise the exact same as  the Apple II version with no new functionality  or customization for the IBM PC specifically.Still, the IBM PC version sold very well for two  years until 1983 when Lotus 1-2-3 was released.  Lotus 1-2-3 was released solely for the IBM PC,  it was developed to run as fast as possible on  it as it solely targeted the IBM PC and was  written in 8088 assembler. It also boasted a  vastly expanded feature set over VisiCalc, whose  sales quickly fell off the proverbial cliff.  Other competitors such as SuperCalc also  played a role in marginalizing VisiCalc,  but it was Lotus 1-2-3 that really  killed its hopes in the marketplace.Additionally a legal battle erupted between  Personal Software and Software Arts that  sucked up much time and money on both sides, and  ensured that little time and attention was given  to developing VisiCalc further. In brief,  Personal Software had come to feel that the  royalty agreement it had signed with Software  Arts, giving them 37.5 percent of every copy  of VisiCalc sold, was far too high, and indeed  it was over twice the 15 percent royalty that  became common once the software industry matured  in the 1980s. However, Bricklin and Frankston  felt that since they had a signed contract for  the 37.5 percent, and both sides were making  quite a bit of money, there was no reason for  them to negotiate with Fylstra over the issue.  This wasn’t the reason for the lawsuits erupting,  but it was an underlying source of stress between  the two companies that exacerbated matters.  Another source of tension between the companies  was Personal Software rebranding itself as  VisiCorp in 1982, and minimizing Software  Arts’ name and connection with VisiCalc,  the product it had created and still owned.Bricklin and Frankston had been working  on a new and improved version of VisiCalc,  called VisiCalc Advanced Version that had a raft  of new features and would have been far more of  a feature competitor to Lotus 1-2-3 on the IBM  PC, had things gone differently. Unfortunately it  was doomed by two poor choices, the first one was  that it wasn’t initially targeted at the IBM PC,  but rather was first released for the ill-fated  Apple III. The second mistake was that when it  was ported to the PC, it was not developed  in 8088 assembler but rather in a high level  language. This made it easier to program, faster  to deliver, and easier to port but also ensured  that it would run far slower than Lotus 1-2-3. The  IBM PC release was also not delivered until 1984,  a year after its original planned release date and  this led to VisiCorp suing Software Arts for 60  million dollars in damages. The bad blood between  both companies was on full display as Software  Arts countersued, stating that VisiCorp had  not properly marketed VisiCalc and furthermore,  owed royalties for other programs that Software  Arts felt were merely extensions of VisiCalc.Eventually the court battles wound up  essentially destroying both companies,  and in 1985 Lotus actually bought both Software  Arts and the rights to VisiCalc. Once Lotus owned  the rights to VisiCalc, it saw no point continuing  development on a competitor to Lotus 1-2-3  and discontinued it entirely in 1985. That  same year also saw Microsoft releasing  the first version of the spreadsheet that  would eventually kill Lotus 1-2-3 in turn,  Excel. Without its flagship product  VisiCorp quickly went defunct and  its remaining assets were sold off  to Paladin Software in the mid-80s.Bricklin went on to form several other tech  based companies and currently is the president  of a software development company called Software  Garden as well as the Chief Technical Officer of  Alpha Software. Frankston went on to work  for Lotus and then Microsoft among others  and today sits on the board of governors  for the IEEE Consumer Electronics Society,  where he writes a regular column for their  magazine. Fylstra had a varied career as well,  and for the past 33 years has been the  president of software company Frontline Systems.And that wraps up the story of VisiCalc,  the world’s first electronic spreadsheet  for personal computers. Future episodes in the  Retro Tech Bytes series are in the pipeline,  including a really neat one that touches the  world of early 2000s Star Wars fan films.  If you enjoyed this video, please hit the  like button, and consider subscribing.