Tech luminaries honor database god Jim Gray
A tribute, not a memorial
In the early nineties, when David Vaskevitch decided that Microsoft should tackle the enterprise database business, the first thing he did was pick up the phone. He dialed the sharpest software minds he knew, and he asked each one who he should talk to about all things DB. They all gave the same answer: Jim Gray. And each described him in exactly the same way: Jim Gray, they said, is smarter than I am.
"I kept going back to the idea that the smartest person in the world must be Jim Gray," says Vaskevitch, now CTO of Microsoft's business platforms division, "because there was no one who would say they were smarter."
Then he met the man. And he realized that if you called him the smartest person in the world, you sold him short. "At least at Microsoft, the smarter you are, the less human you are. But Jim was exactly the opposite."
In 1998, Jim Gray won the Turing Award for his seminal work on database transaction processing. While at IBM's San Jose research center in the mid-70s, he helped define the fundamental ways that information moves to and from the world's online record stores. But for Vaskevitch - and for so many others - his influence was much greater than that.
"A lot of the core concepts that we take for granted in the database industry - and even more broadly in the computer industry - are concepts that Jim helped to create," Vaskevitch says, "But I really don't think that's his main contribution."
Jim Gray didn't just nurture the database. He nurtured the database industry - and industries beyond - serving as friend and mentor to an army of pioneering computer scientists. "I would call someone up and try to convince them to come to Microsoft," Vaskevitch remembers, "and sometimes, they would say they wanted to talk to Jim Gray first. But usually, they had already talked to Jim about the phone call I was going to make before I even knew I was going to make it.
"Jim had this web of connections. You can think of him as the transaction coordinator, where the transactions were people moving around the industry. He made it all work."
And making it all work wasn't his motivation. "Jim had an ability to make you feel that he really cared about your life - because he did - an ability to have that kind of personal relationship with so many of us. That's what most defined him."
On Saturday, at University of California Berkeley, David Vaskevitch was among the dozens who gathered for a day-long tribute to the life and work of Gray. Eighteen months ago, Gray was lost at sea after sailing his yacht, Tenacious, from San Francisco Bay.
Despite a five-day hunt by the US Coast Guard - and another three weeks of searching by some of the tech industry's cleverest minds - he wasn't found. But in the words of Pauline Boss, Saturday's gathering was not a memorial.
"It is because of the mystery that we honor rather than memorialize Jim Gray today," said Boss, professor emeritus of psychotherapy at the University of Minnesota. "Paradoxically, acknowledging what we don't know helps focus on what we do know: that Jim Gray's contributions to the world, to science, to friends, and to his family are immense, and continue to influence us all every day."
The tribute detailed Gray's entire career from his days as the first computer science PhD at Berkeley, to his stints at IBM, Tandem, and DEC, to his most recent work at Microsoft's Bay Area research lab. And each speaker - including Microsoft Research founder Rick Rachid; DEC/Microsoft legend C. Gordon Bell; and Ingres founder Mike Stonebraker - painted the database pioneer as one of the industry's great minds. But in a constant echo of those words from Vaskevitch, praise for Gray the computer scientist was inevitably dwarfed by praise for Gray the man.
"Over the years, I would frequently talk about Microsoft Research and about Jim and about our research lab in the Bay Area, and I always said the same thing," Rachid told an audience of several hundred at Berkeley's Zellerbach Hall. "I would say 'Jim is the kind of person you'd want your child to be like.'
"We all want to see our children grow up to accomplish something. We want to see them have an impact on the world. But we also want to see the exemplify human values. We want to see them be kind and generous and thoughtful of others.
"Jim was incredibly smart, incredibly accomplished. But at the same time, he was thoughtful. He was an unassuming person. He was incredibly generous. I think it says something about Jim that in almost all the conversations I remember having with him, he was talking about someone else. He was talking about someone else's accomplishments and someone else's problems and what we could do to solve them."
To be sure, it's no small feat that the man overshadows the researcher. In the words of Bruce Lindsay - who worked alongside Gray on the "System R" project at IBM Research - Gray's transaction processing work is "the key to life as we know it." And this was only a slight exaggeration.
Gray's famous criteria for online transactions - atomicity, consistency, isolation, and durability (ACID) - underpin everything from online banking, booking, and billing to the so-called war on terror. "The No-Fly List," Lindsay said. "Brought to you by Jim Gray." Those fundamental concepts ensure that modern databases play well with one another - and typically recover from hardware failure.
"Jim's major contribution was to performance and efficiency," Mike Stonebraker explained. "He figured out fast mechanisms to guarantee [reliable databases]." In 1976, Gray co-wrote a paper with other IBMers that first defined these ubiquitous database algorithms, and in 1992, together with German researcher Andreas Reuter, he published the field's 1,124-page magnum opus: "Transaction Processing: Concepts and Techniques.
And that was just a start. Jim Gray joined Microsoft Research in 1995 - after Rachid opened that Bay Area outpost just for him - and it was there he helped launch TerraServer, a massive online repository of aerial and satellite imagery. Once again, he was ahead of the curve. TerraServer would eventually give rise to Microsoft Virtual Earth.
"I always think into the future, but no matter how Jim was always one step ahead of me," said Vaskevitch. "We wanted to build a really, really, really big database and project it out onto the web, and he decided to build it around satellite imagery that would let you see the earth's entire surface.
"At the time, I thought this was - I wouldn't say crazy, but I would say not all that relevant. Of course, today, everyone is used to the fact that you can sit down - even with a cell phone - and look down on a picture of your own house."
Such technology would also drive the search for Jim Gray and Tenacious. On the Thursday following his Sunday disappearance, the Coast Guard finally called off its boats and planes, which had combed over 132,000-square-miles of ocean. But the hunt was continued by literally thousands of friends and admirers - volunteers who who spent the next three weeks poring over satellite images of the waters off San Francisco Bay.
But as ex-Oracle VP Mike Olson explained at Saturday's tribute, this effort exceeded the bounds of TerraServer, Virtual Earth, and Google Earth, which rely so heavily on older images. "Satellite imagery is hard to get and hard to use," said Olson, who led the amateur search. "But Jim's high-powered group of friends included people who could re-target satellites, who could move them into new orbits and have them look down on new areas."
The search wasn't successful. But it too was a tribute to Jim Gray, painting much the same picture that Vaskevitch's paints. "Those three weeks were among the most intense and meaningful of my personal and professional life," Olson said. "I didn't like it. But I wouldn't have missed it for the world."
Video of the Jim Gray Tribute will soon be available here. ®