Computing gets personal - The Computing Universe: A Journey through a Revolution - page 169

Lampson was a graduate student and they had collaborated on the CAL TSS time-sharing system. He had also
worked with Lampson at the ill-fated Berkeley Computer Corporation while still a student. After working
for a while on a parallel computing project called Illiac-IV, he rejoined his ex-Berkeley colleagues at PARC.
Lampson supplied Simonyi with three sheets of notes capturing his thoughts for an interactive text editor.
Simonyi called his new word processing system “Bravo.” Using the Alto's bitmapped screen, he was able to
encode complex typefaces, boldface, italic, and underlining in the text together with a detailed page layout,
so that the document appeared on the screen almost exactly as it would be printed out. Bravo was thus the
first WSYIWYG word processor - What You See Is What You Get - and it became a great hit among the engi-
neers at PARC. As Simonyi later said:
It was the killer app, no question. People would come into PARC at night to write all kinds of stuff, sending
letters, doing all personal correspondence, PTA reports, silly little newsletters, anything. If you went around
and looked at what the Altos were doing, they were all in Bravo. 42
In spite of its popularity with the engineers at PARC, Bravo needed a more user-friendly interface if it was
to be adopted by the much larger community of nonengineers. Lampson and Thacker had made a deliber-
ate decision not to work on the user interface of Bravo, not because they did not think it was important but
because they did not have the resources to do both the implementation and the user interface. It was left to
Bob Taylor to initiate such a project with two other computer scientists at PARC, Larry Tesler and Tim Mott.
Before joining PARC, Tesler had produced a program called Pub that helped ordinary users format and
print their documents. At PARC, Tesler had been a member of a team trying to reengineer and update a ver-
sion of Engelbart's interactive multimedia system. He rapidly became dissatisfied with the complexity of the
system being created and was eager to take on a new challenge. Tim Mott was an Englishman with a computer
science degree from the University of Manchester who was working in the United States for a Xerox subsidiary
called Ginn & Company that published textbooks. Determined to try to get some value from Xerox's “corpo-
rate research” tax, Mott's boss Darwin Newton sent Mott to visit PARC and see how their research on office
systems could assist him as a publisher. Mott concluded that their system was much too complex and difficult
for the publishing company to use: “There wasn't a lot of time spent looking at what mere mortals would be
able to do with the system.” 43 Taylor challenged Mott to use the Alto to produce something useful.
Tesler and Mott also found the user interface of Bravo far too complicated. It was usable by experts but not
easy or attractive for ordinary users like publishers. Mott went back to Ginn & Company and did some market
research on what nonengineers actually wanted from such a program. Unsurprisingly, he found that the pub-
lishers wanted the program to mimic what they actually did with their paper-based process. This is the origin of
the “cut” and “paste” commands that are still used to this day. Tesler and Mott called their new system Gypsy,
and it was the first program to use the mouse to execute point-and-click operations in the way we do today.
While Simonyi, Tesler, and Mott were developing Bravo and Gypsy, Alan Kay's group at PARC was still
pursuing his Dynabook dream. The Alto's bitmapped screen allowed enormous flexibility in what could be
displayed. So why can't a user write a memo in one part of the screen and use a drawing program in another
part? This led the team to think of the screen in terms of a “desktop” metaphor where electronic docu-
ments could be piled on top of one another, just like papers on a desktop. They created overlapping boxes,
or “windows,” for each different task. But shifting these boxes around put a huge demand on the Alto's
processor and was extremely slow. In a stroke of genius, Dan Ingalls came up with “BitBlt,” an abbreviation
of “bit boundary block transfer” and pronounced bitblit . Instead of having the computer change each of the
components of a rectangular image individually for the new location, BitBlt operated on the entire bitmap
using fast Boolean operations to create the new image. This new technique meant that the user could rap-
idly scroll up or down the text of a document on the screen by moving a mouse. It also meant that windows
could be created and moved around at will, and that the illusion of a stack of papers on a desk could be