Very few made it to market because it was something of a Darwinian development environment at Apple. Out of 40 product concepts maybe then 5 get prototyped and maybe 1 got to market. We did make 200 hundred or so of the Toshiba touchscreen CDROM book device but we had issues with bringing the battery life and the price point together in a package that made sense.
The work had led him to see a key flaw in the Mac OS.
In a sense the catalyst for change at Apple came from the fact that the time-based concept called multimedia was becoming popular and it should have been a natural area for Apple to be in but we had nothing. John Sculley told me, “We need to get into this”, but of course it wasn’t that easy. I started looking at how we could leverage the company’s multimedia potential.
The Mac was extremely popular as a graphics layout and publishing tool but the OS had been built quite a few years earlier with the paper/folder/file metaphor in mind and as a result there was nothing to account for time-based events in media.
There were crude timers that just measured cursors moving across but nothing for data. There was no way to start unless we had a way to address time based events in the operating system, so we had to address a pretty sophisticated challenge from day one.
Andrew Soderberg, later a QuickTime hardware evangelist:
For good reason everyone thinks of QuickTime as being all about video and audio, but from the start of development, it was all about temporal data. Events happening in time, in a sequence. It could just as easily have been famous for the underlying technology to home security systems.
Managing when things went on or off. Like HomeKit is today. Or Notifications on your iPhone is today. In QuickTime you can have multiple ‘tracks’, and they can all do their own thing as long as they are all in this one temporal QuickTime container.
Everyone now imagines ‘tracks’ as meaning video or audio tracks but that wasn’t necessarily the use case behind the development. It really didn’t matter ‘what’ was on the track, as the software came along and saw an event on the track, it reacted as programmed at a point in time. It was a ‘document’ over time, with no start and no end.
But as Tyler says, this was going to be a big deal. A big deal to change for Apple. And proof of that was when we later released QuickTime for Windows, it wasn’t just a plug-in that was added to Microsoft’s operating system, we practically created another time based environment inside Windows to allow QuickTime to run.
Peppel created a product specification for an extension to the Mac OS to handle time based multimedia elements like video, graphics and audio. Middle management’s disdain for change echoed that of Ampex in the 1950s. Charlie Ginsburg’s development of Ampex’s videotape machine was stalled or stopped several times in favour of the company's existing audio products.
Peppel’s time based software idea was put to the side. For the sake of hardware. Within Apple’s in-house Advanced Technology Group (ATG), Steve Perlman was working on hardware supported video compression. Perlman christened his hardware compression device, QuickScan and John Sculley showed an interest:
We had a very talented engineer named Steve Perlman, who also having challenges getting along with the mainstream engineers and so I moved Steve over to my office and so he sat outside my cube. Why was this all significant to me? Because to me it was just another step along the way to the Knowledge Navigator.
Duncan Kennedy recalls:
For about three to five years there was a major effort within Apple to develop hardware based video projects.There were projects trying to leverage HyperCard or projects like QuickScan but no one was really doing software based video.
Around the same time, Bruce Leak graduated from Stanford and applied for a job at Apple Computer.
I had 4 interviews, got turned down in three of them but finally found a job in System Software. It was a dream job for me. The Macintosh was the most interesting computer out in the world.
EDUCOM
John Sculley delivered the keynote at EDUCOM 1987 in Los Angeles. After an introduction he played a short video called The Knowledge Navigator, which showed a real life scenario between a University professor and a device that could access a large networked database of hypertext information. It used a software agent, depicted as a bow tie wearing assistant, to help in searching for information.
The vision brought together a wish list of computer experiences that was unbound by Moore’s Law or contemporary technology. David Bunnell called it
“…something between a product announcement and a sci-fi story.”
Sculley later told Laurie Flynn at Infoworld:
This tool is one that really has all the key technologies underway in the 1980s that we’re going to need by the end of the century.
While it impressed the audience, the reception elsewhere was measured. The product video was not widely reported until Sculley repeated the demonstration of Knowledge Navigator at MacWorld in early 1988. This time the video played to Apple devotees, developers and journalists and it seemingly set Apple apart with a clear vision for its future.
Alan Kay later told Mark Brownstein:
The Knowledge Navigator is John’s version of the Dynabook. It’s been inspirational. Everyody at Apple us thrilled that John is supporting a concept. ...it helped to focus people, it became more real. We’re getting the company to start thinking about systems. The systems will empower the individual through things that are networked.
The concept video also reminded everyone of a technology conundrum. What could be imagined was often not able to be built. Apple's Group Product manager Tyler Peppel cautioned in an interview for InfoWorld:
Knowledge Navigator is not a product plan but a vision of where we could go.
Jean-Louis Gassee was Apple’s President of Products:
Multimedia is not poised for success like desktop publishing was in 1985. Those who think otherwise will be in for a major disappointment. I have every confidence that the required hardware building blocks will eventually become available and affordable, but the transition from desire to reality will be much slower than in the case of DTP.
Pundit Jonathan Seybold cautioned in the LA Times.
People habitually overestimate the speed with which something happens in the short run and underestimate the impact in the long run.
In future, the Knowledge Navigator became an Apple product suited to many professionals including editors. A handheld device able to edit graphics, multitrack audio, broadcast quality video and real time special effects capabilities. But an iPad with Final Cut X and Siri was more than twenty years away.
STOLEN GOODS
Fifteen months of work culminated in October 1987 with the launch of Digital F/X’s first shipping product. Jason Danielson recalls the technical preview of the DF/X 200 Production System at SMPTE in Los Angeles:
We were scrambling to get all the code written and complete before we opened the booth which had a banner that said we were “A Paintbox and ADO but in a single box". The trouble was then that the system kept bugging out so we decided to take the banner down.
With its modern GUI and digital throughput the DF/X 200 signalled where the online postproduction industry was going. Here was a single unit capable of compositing and special effects without any image degradation. Danielson