SIMSON SAYS: Copyright and wrongs

Simson L. Garfinkel (
Thu, 25 Feb 1999 10:55:32 -0800

A little more than a decade ago, Lotus reinvented the spreadsheet. A small
group of developers at the company's Advanced Technology Group, including
Pito Salas, Paul Kleppner, and Lynda Urgotis, started with the mission of
making spreadsheets dramatically easier to use. A big problem that many
people having building complicated spreadsheets, they realizes, is laying
things out. What should go in the rows? What should go in the columns?
Spreadsheets like Microsoft Excel and Lotus 1-2-3 allow users to put any
number or any formula that they want in any cell on a huge 2-dimensional
array. Unfortunately, for most users, that flexibility provides more power
than is needed, and creates a lot of confusion in the process.

A few months after moving into an empty library on the first floor of 161
First Street in Cambridge, the group had come up with a better way. Instead
of making the user type seemingly meaningless formulas like C1=A1-B1, and
then having to replicate these formulas throughout the spreadsheet, the
group created a prototype that let the user type meaningful formulas, like
"PROFIT=PRICE-COST." Many cells could then be labeled as PROFIT, PRICE, or
COST cells, and the formula would automatically be applied to each one.
Users might create separate PROFIT, PRICE, and COST cells for each year that
they were analyzing. Or they might have one for each region of the country.
Or --- and this was the really radical idea --- they might have different
PROFIT, PRICE and COST cells for different regions and for different years,
creating a three dimensional spreadsheet. The user might then add a fourth
dimension, to investigate different business scenarios, and even a fifth
dimension, to look at different products.

Pitlo Salas took a small multi-media demo of his "Modeler" program on tour
to some of Lotus' largest accounts in the spring of 1987 and got rave
reviews. In September 1988 the company formally committed to the project,
code-named "Back Bay." The following month Steve Jobs visited the company,
and convinced Lotus Management to put Back Bay on Job's brand new computer,
a black magnesium machine called the NeXTcube. The product made its debut in
February 1991. It was called Lotus Improv.

At this point, the story is better known. Jobs didn't do too well with NeXT:
he never sold more than 100,000 of the black computers. So in 1992 Lotus
ported Improv to Microsoft Windows 3.1. But it didn't do too well there,
either. According to Kate Monaghan, a spokesperson for Lotus, " Improv won
numerous awards as an innovative spreadsheet tool but was discontinued in
April 1996 due to low sales. Customers had a hard time deciding between
1-2-3 or Improv and so stuck with 1-2-3, the industry standard."

Today the industry standard is no longer 1-2-3: it's Microsoft Excel. And
while Excel has a few of the features of Improv, it lacks the fundamental
core. I didn't really think much about Excel's shortcomings until a few
months ago, when I started trying to build a series of extremely complicated
Excel spreadsheets. I was doing some financial projections for a friend's
company. I wanted to chart the sales of several products over time,
exploring different economic and pricing scenarios. Then we wanted to look
at different funding scenarios. That's five dimensions right there. Excel
has a feature called "pivot tables," but it really only works with two
dimensions. You can also stack multiple Excel spreadsheets together to
create a third dimension. But after that, Excel is rather helpless.

I am one of the few people who had a NeXT back in 1992, and I got to be
pretty skilled at using Improv. So after spinning my wheels with Excel and
buying more than a hundred dollars with of books that promised to solve my
problems, but delivered far less, I started hunting on the Internet. My Holy
Grail was a 1993 copy of Lotus Improv 2.0 for Windows 3.1. After a lot of
searching I found somebody who had an old copy, and I bought it for $75. I'm
pleased to report that the program runs like a champ under Windows 98. And
it's remarkably small: the entire program, including the help files, the
multi-media tutorial, and the example models, takes just 9 megabytes of
space on my hard drive. I haven't given up on Excel, but quite frankly I
haven't found a need for it since switching.

Unfortunately, that's the end of the story. I'm legally prohibited from
making copies of Improv for my friends and co-workers. Improv is protect by
copyright, and even though Lotus no longer sells or supports the product,
that protection still holds. In a very real way, this is a kind of crime
against society. The purpose of copyright under our legal tradition is to
give authors protection for their intellectual works, so that they will be
motivated to produce new ones. It is a bargain between society and the
inventor. If a company like Lotus reneges on its side of the bargain --- and
in the case of Improv, Lotus surely has --- I think that the works that were
protected by copyright should become public property. That is, I think that
they should go into the public domain.

The implications of this argument go far beyond Lotus and Improv. In recent
years dozens of software publishers have gone out of business, and taken
their wares with them. Just imagine what the world would be like today, if
instead of killing their products, these companies had been forced to
release their programs into the public domain. Today there would be more
than a dozen free word processors and spreadsheets available for Windows,
giving Microsoft a real run for its money. And if the companies had been
compelled to release the source code for these products as well, then
enterprising hobbyists would have ported these applications the Linux
operating system.

One of the prime offenders in this world of dead software is Apple, which
has mothballed both the Newton and all but given up on the NeXTSTEP
operating systems. Ultimately, it would probably be in the best interest of
both Apple's shareholders and society as a whole for companies like Lotus
and Improv to release their failed products to the public. There is a simple
reason that they do not: if the public did a better job with the software,
then it would prove that these products died because of mismanagement, and
not because of the competitive environment.

The story of Lotus Improv can be found at . A comparison of Improv with a
conventional spreadsheet can be found at

[Screen shot at] Cutline: "Lotus Improv is
a multi-dimensional spreadsheet that seperates data (top) from formuals
(bottom). It's a great program. Pitty you can't buy it."

This message (C) Simson L. Garfinkel. .
This message was sent to you by way of the SIMSON-SAYS mailing list.
SIMSON-SAYS is Simson's column on computer issues that appears weekly
in The Boston Globe and other newspapers.
This column is available for syndication. If you would like to have
SIMSON-SAYS appear in your local newspaper, tell them to drop me a
Please feel free to pass this e-mail on to a friend. You can subscribe
to SIMSON-SAYS by sending the message "subscribe simson-says" to the
address "". You can take yourself off the
simson-says mailing list, by sending the message "unsubscribe
simson-says" to the same address.