John Buck at The Verge (gift link), excerpted from his great book, Inventing the Future:
Steve Perlman: Almost everyone at Apple, and definitely
everywhere else, assumed that multimedia would always require
specialized hardware — and be expensive. A few of us thought
otherwise.One of the few was Gavin Miller, a research scientist in Apple’s
Graphics Group, who worked with Hoffert to crack the problem of
software compression and decompression, otherwise known as codec.Gavin Miller, research scientist: We went for a lunchtime
walk, and by the end of it, we had generalized the model to
include constant color blocks and 2-bit per-pixel interpolating
blocks. This allowed us to trade off quantization artifacts in
large flat areas for more detail in textured areas. The result was
an increase in quality and performance that helped to make the
codec practical for really small video sizes.
Just a typical lunchtime walk-and-talk.
Fun anecdote from 1990:
He asked Peppel to create a product plan that he could announce at
Apple’s Worldwide Developers Conference on May 7th. That day,
Casey took to the stage and announced QuickTime to a stunned
audience, saying, “Apple intends to develop real-time software
compression/decompression technology that will run on today’s
modular Macintosh systems. A system-wide time coding to allow
synchronization of sound, animation, and other time-critical
processes.”Casey explained that Apple’s new multimedia architecture would be
delivered by the end of the year. He did not say that QuickTime
had no budget, staff, or offices.Worthington: We were dumbfounded.
Konstantin Othmer, QuickDraw engineer: I was standing next to
Bruce Leak, and asked him, “What the heck was that?” He said he
had no idea.
QuickTime actually shipped by WWDC 1991, teaching Apple the important lesson that anything they announce at WWDC, no matter how premature, will ship as promised.
