Whatever Happened to Context MBA?
by John C. Dvorak
It’s easy to forget that it was the microcomputer industry that gave birth to the IBM PC and not the reverse. Or that Lotus 1-2-3 didn’t always own the spreadsheet market. A strange line in an old review of Context MBA reminded me of this. The reviewer, reflecting on the “powerful” IBM PC that had been “loaded” with 384K of RAM, wrote, “OK, here we are with this big PC just ticking over like a 1964 V-8. Now what?” Even at the time, this must have sounded corny, yet in 1982 an unopened box of software still held the promise of something completely new, something that hadn’t been done before. Context MBA was one of the few programs to fulfill that promise.
MBA was the first integrated software package. At $695, it, like most of the other early integrated packages, was targeted at the high end of the market. Although we look back on MBA as an integrated package, when it was released in mid-1982, there were no such convenient pigeon-holdes. One magazine even called it an “Omnibus” program when it first appeared, which couldn’t have helped its marketability. As a result, MBA was usually compared to “financial modeling” or “electronic” spreadsheet programs.
VisiCalc, written by Dan Bricklin and Bob Frankston, was the first successful spreadsheet program. Within months of its release in 1979, on the Apple II, it topped all the software sales charts. It also made the Apple II important for business users. At the same time, however, Personal Software, which later changed its name to VisiCorp, neglected the CP/M market, allowing Sorcim to dominate with a VisiCalc knock-off called SuperCalc. As VisiCalc sales began soaring towards the one million, rumors were circulating that IBM would soon enter the microcomputer market.
MBA debuted on the ill-fated Apple III. Apart from the bad luck of having chosen a doomed platform, MBA successfully combined financial modeling, business graphics, relational database management, and word processing functions into one package. This was a major first for microcomputers because, prior to the introduction of the IBM PC, most applications’ programmers had to fit their program’s operations into what now seems like a ridiculously small amount of memory–32K of RAM or so. Even in the days of supertight coding, 32K was barely enough space to carry out one program’s operations, much less those of four or more.
IBM ENTERS THE GAME
At the same time, IBM was busy lining up software for its new machine, the PC, behind a heavy veil of security and non-disclosure statements. Although few remember it now, Microsoft’s Disk Operating System was not the only operating system to receive the official Big Blue stamp of approval. When IBM introduced the PC in August 1981, they promised that in addition to Microsoft’s DOS, which shipped with each PC, both Digital Research’s CP/M 86 and the UCSD p-System would be available. However, Microsoft’s rivals were both so late that MS-DOS was never seriously challenged.
With VisiCalc’s dramatic impact on the sales of Apple II’s to business users fresh in their minds, IBM executives were careful to display a demonstration of the PC version of VisiCalc at the PC’s unveiling. That, together with VisiCalc’s established reputation as the leader in spreadsheet software, propelled it to the top of the PC software sales charts. Meanwhile, Context MBA 1.2 for the IBM PC didn’t leave the shipping dock until May 1982.
When it did, users hailed MBA’s features. It allowed managers to quickly and easily prepare reports, graphs, and spreadsheets that otherwise took days to prepare by hand. MBA also allowed the user to open separate proto-windows for different spreadsheet, graph, database, or word processing files.
One of the unique features of Context MBA was that a single cell, in the so-called modeling context, was capable of holding approximately four pages of text, as opposed to the normal 256 characters. That, combined with features such as the telecommunications “context” and the simple macro facility that were added in later versions, allowed the user to store a memo, a graph, a communications script, and a macro in separate cells of a spreadsheet. With all this under one roof, MBA made it possible to automate common computing tasks.
FOR WHOM THE BELL TOLLS
Unfortunately, Context MBA was written in the UCSD Pascal p-System. At a time when portability was at a premium because of the variety of computer sytems in use, the ease with which the p-System allowed developers to port their programs from one system to the next must have seemed attractive. Unfortunately, as we saw in an early “What Ever Happened To?” column, the p-System’s portability came at the expense of performance. Screen rewrites were slow. Scrolling was slower. And with large spreadsheets, recalculations could last minutes!
Context Management Systems President G.H. Hoaxie argued that even with the slow performance, MBA’s other efficiencies more than made up for its sluggishness. But once spreadsheet users felt the relative performance snap of VisiCalc or SuperCalc, they were unwilling to give it up, regardless of whatever else they might gain in the switch.
MBA had other problems, too. For example, the early versions of the program used an odd copy-protection scheme that caused frequent, and annoying, disk accesses. Another problem was limited distribution. Dealers were required to take a 3-hour, 30-minute training session on the software and were then made to pass a test demonstrating their understanding of the program. While admittedly not a bad idea, especially to anyone who has ever suffered at the hands of ignorant dealers, in practice this policy meant that the original IBM version of MBA was available from only 125 dealers nationwide. On top of all that, it took users between eight and 24 hours of training to learn to use the program.
Meanwhile, in January 1983, Lotus 1-2-3 began shipping. 1-2-3 release 1A was similar to VisiCalc in operation. It also included most of the graphing and database capabilities of MBA, but it was optimized for the PC. 1-2-3’s speed, in particular, was superior to both VisiCalc and MBA. Since the IBM PC version of VisiCalc was basically the same as the Apple II version, VisiCalc was vulnerable to any spreadsheet that capitalized on the PC’s relative advantages over earlier microcomputers.
Once again, VisiCorp ignored a lucrative market until it was too late. By the time they released VisiCalc Advanced Version, opportunity had slipped away. Three months later, 1-2-3 knocked VisiCalc out of the number one slot it had held in software sales for the previous four years. And Context MBA was buried.
Although Context MBA didn’t succeed as either a spreadsheet or an integrated package, it survived for three or four years. Its features and speed were continually improved, and eventually the entire program was renamed Corporate MBA and rewritten to run under DOS. Nevertheless, the bell had tolled loud and clear.
SON OF MBA
In the computing equivalent of passing genes down the ladder of time, Context MBA left several indirect legacies. The most obvious was its combination all of the major functions that a computer user needs into one package. MBA was different from 1-2-3 in this respect, because 1-2-3 was first and foremost a spreadsheet program. Interestingly, Jonathan Sachs, the author of 1-2-3, decided against including a word processing function after seeing a prototype of MBA. Sachs believed that the word processor contributed to MBA’s poor performance. In the days of 4.77MHz 8088s, performance was a major factory on every developer’s mind.
The other major legacy was Robert Carr. He was one of the contract programmers who worked on the first implementation of MBA. His experiences with the program led him to develop a similar program named Framework for Ashton-Tate and then went on to head of development at Go Corp., which developed a pen-based, object-oriented operating system. After that he was involved in some of the first web-based e-commerce sites.