Dr. Dobb's Journal October 1997
A few months ago, I added a news page to my web site, Swaine's World (http://gate.cruzio.com/~mswaine/ swainesWorld.html). I'm not sure why I need a news page; there are plenty of good ones out there, and all I do is summarize and comment on news stories published elsewhere. I guess it's a handy way to force myself to stay on top of a lot of developing stories from the news sources I follow, including the ongoing Apple Computer soap opera. My news page has an archive, too, or maybe I should say a morgue (the news page is at http:// gate.cruzio.com/~mswaine/grapevine.html and the morgue is at http://gate.cruzio .com/~mswaine/grapecellar.html). Back in July I posted the following Apple news to my news page:
July 10, 1997: Apple CEO Gil Amelio has resigned. Apple issued a press release yesterday announcing the resignations of Amelio and Senior Vice President of Technology Ellen Hancock. A search has been initiated for a new CEO. The search team consists of four people: Steve Jobs, Mike Markkula, CFO Fred Anderson, and one other board member, Edgar S. Woolard, Jr. The press release spoke of a need for a "customer focused" CEO.
While the search is on, Anderson will take on an expanded operations role, working with the board and the executive management team, while Jobs will take on an expanded advisory role, advising the board and the executive management team.
After I posted this story, Apple announced its third fiscal quarter results on Wednesday, July 16. The company only lost $56 million, not as bad as expected. Predictably, the San Jose Mercury News spun this positively, the Wall Street Journal characterized it as less bad than expected, and PC Week gave it a negative spin. CFO Fred Anderson said that Apple was going to stop trying to predict publicly when it would be back to profitability. Good plan.
The news items kept flashing from Apple in the days that followed -- new machines, price cuts, the hired headhunting firm (Havoc & Struggles, was it? Oh, sorry -- Heidrick & Struggles), and the release of OS 8. I tried to keep up with all of them, and maybe I did, but my news page missed a few.
Probably by the time you read this Apple will either have hired Ross Perot as CEO or Michael Jackson will have bought the company, but here is my Monday morning view of that July news flurry from the company that has red ink flowing at home and black ink flowing in every pressroom: Amelio was fired. Rumors had been circulating for days that he was about to be let go. Arguments circulated for weeks afterward about whether he quit or was fired. It was pretty obvious that it wasn't his idea. Ellen Hancock, on the other hand, may have resigned, not that it made any difference. Hancock had already been effectively demoted in the wake of the NeXT acquisition and subsequent restructuring, and her former job had been divided among Jobs' protégés John Rubenstein, senior vice president of hardware engineering, and Avie Tevanian, senior vice president of software engineering. Some insiders say that she was responsible for keeping a strong focus on OS 8 while Rhapsody is being developed and established. I wondered what she was doing. But whatever role she still had after NeXTification has now been absorbed by Rubenstein and Tevanian.
Rubenstein and Tevanian came to Apple from NeXT. The fact that they were immediately placed in charge of Apple's hardware and software direction was a clear signal that NeXT technology was going to become Apple's focus and that Jobs's vision for the company was going to be pushed hard. Even without Jobs acting in an advisory role, this was assured. Tevanian and Rubenstein have the Jobs vision in their DNA. I've heard Steve Jobs's words come out of Avie Tevanian's mouth (or possibly vice versa).
In the months following the NeXT acquisition, all signs have indicated that, sure enough, the Jobs agenda is the rule of the day. Rubenstein and Tevanian seemed to move ahead with their agendas without obstruction, and one after another of the opinions that Jobs expressed publicly or privately became Apple policy, from personnel changes like the shunting aside of Hancock to technology decisions like killing off OpenDoc and setting up the Newton division for sale.
Then there's the CEO search team. Markkula is Apple's largest shareholder and has been on the board for 20 years. He has had a hand in the hiring or firing of every Apple CEO, and some say was the ultimate decision maker in every case. He stripped Jobs of power a decade ago, causing Jobs to leave the company and start NeXT, and he certainly approved, if not engineered, Jobs's return as "advisor" when the NeXT deal went down. Whatever rancor existed between the two has apparently healed. Given that Anderson's role in the CEO search is likely to be limited to operational and financial considerations, the selection of Amelio's successor is effectively in the hands of Steve Jobs and Mike Markkula. And, given the personalities and personal agendas of those two individuals, that probably means a Steve Jobs selection. The incredible coup that began last December is now complete. Steve Jobs is calling the shots at Apple Computer.
That same day in the San Jose Mercury News, Dan Gilmore's column on the Amelio departure puzzled me. "For a tiny, shining moment," he said, with the NeXT acquisition, there seemed to be a plan. "Now, again, it's all up in the air." But the consensus is that since the NeXT acquisition Amelio has been disengaged from the company and that Jobs has been defining the agenda. That's certainly how it looks to me. So why should the plan be up in the air? If anything, I suspect that things may move faster without a CEO. Where they'll move and whether that'll be what's needed are other questions, of course. The soap opera continues, and it matters, but it seems to me that the plot of this particular installment is not "Amelio forced out" but rather "Jobs consolidates position."
I trust that Hancock's wasn't the only voice defending the MacOS within Apple. Apple has to show strong commitment to the MacOS for as long as it takes Rhapsody to catch on and beyond that.
Apple released OS 8 in late July to favorable but confusing reviews. PC Week said it was the biggest upgrade in six years, PR Newswire accepted Apple's spin that this was "the most significant MacOS upgrade since 1984," and the Merc saw it as no big deal.
The release of OS 8 is a very big deal for Apple. But it isn't so important that this version be the biggest thing since 1984, as Apple foolishly boasts. What is important is that it's a significant improvement, that it works, that it offers enough advantage so that a significant number of people upgrade and are not disappointed when they do, and that it arrived on time.
I don't know how many people will upgrade, but Apple has hit all the other bases. The OS adds a new look, is apparently more stable, marginally faster, has a multithreaded Finder, and has impressive integration of the Internet and Java on the desktop; but it still doesn't have memory protection or preemptive multitasking.
But that's enough. Apple needs to project an image of a company that can deliver good-enough software on schedule. If I were to offer Apple a slogan for the rest of 1997, it would be Adam Osborne's "Adequacy is sufficient." Next year is soon enough for a return to claims of Insane Greatness.
Apple killed off OpenDoc this year. Apple boosters hate it when you say it that bluntly, but how else do you interpret "we put a bullet in its head," which is exactly the phrase I've heard from both Steve Jobs and Avie Tevanian?
But if it is dead, it now rises from the grave. Apple has granted Hutchings Software (http://www.hutchings-software .com/; or e-mail Brad Hutchings at brad@hutchings-software.com) a source-code license for ODF, the OpenDoc Development Framework created by Apple for creating OpenDoc part editors. Brad has since released two new versions of ODF, freely available to any developers who want to use them.
"OpenDoc is back," Brad exults. To stay abreast of user component developments on the OpenDoc paradigm, you can join the odf-dev mailing list at http://athena .minerva.net:8081/guest/RemoteListSummary/odf_dev or visit the home page at http://www.componentx.com/odf.
Componentx? Right: The "Live Objects" label for OpenDoc components is no longer available. The Hutchings Software brand label for components is "Component X."
Apple commentary is getting to be a dangerous pastime. Say something critical of Apple and you're likely to be challenged, threatened, or spammed.
Some of the culprits are Apple supporters who subscribe to certain mailing lists that I won't name. I call them "spin vigilantes," and I mean that in the nicest possible way. I read their posts, and I know that they are a source of good information when they are not beyond their areas of expertise, like planning Apple's strategy.
Still, it hardly seems fair for MacUser's Rik Myslewski to characterize them as chimpanzees. "A bunch of potbellied, has-been sandlot ballplayers, sitting around a sticky table at their local watering hole, nursing warm beers" is another way that he characterizes anyone with the effrontery to publish their views on What Apple Ought To Do.
You would think that spin vigilantes with time on their hands would be digging through back issues of MacUser for examples of Myslewski-bylined advice to Apple. Instead, they're quoting Rik's column with relish. Why? Because they think Rik's real target was someone else. Dave Winer thinks it was him.
Winer, who comments almost daily on Apple events at http://www.scripting.com/, is getting it from all directions. Usually seen as a supporter of Apple and its technology, he has written some more critical pieces this year, and the fans felt betrayed. They said so, and then Rik M. had his say, and pretty soon Winer was writing conciliatory pieces and doing his best to lower the volume (in two senses) of his incoming e-mail.
Dave did a good job of smoothing things over, but this attacking of journalists could get out of hand (like, somebody could spam me). Stewart Alsop thinks it's already out of hand. Writing in Fortune, he quoted one response to his criticism of Apple: "Many Mac fans that made the smart choice of buying an Apple product will have discussions about you and your death."
This is of course atrocious. It should be "Mac fans who", not "Mac fans that."
I'm sure that Dave lets no conscious bias influence his writing. I've been reading him forever and I consider him the best source of news analysis on all matters Macintosh, period. But I also know when to take his views with a grain of salt. When he obsesses over Apple's treatment of developers, it's a developer talking. When he encourages Apple not to compete with third-party products, he's thinking of what Apple did to him. And when he campaigns for the MacOS rather than Rhapsody, he's thinking of his investment. A truly unbiased voice, if there were one, would acknowledge that if Apple has a future, it's Rhapsody. As for Stewart, comparing Guy Kawasaki to Joe McCarthy and hinting darkly that it's time we developed laws and civilized cyberspace seem to me like overkill, but of course I'm not the one getting death threats. It does puzzle me that Stewart published his response to his attackers in Fortune, a publication that he clearly thinks they don't read.
I recently had a chance to see a disassembled Illiac-IV, a piece of computer history (and a large piece it was), at NASA's Ames Research Center in Mountain View.
Ames had its own Illiac-IV from 1972 to 1982. When it was delivered, it took three years to get it online. When it was decommissioned, I was living just up 101 in Whiskey Gulch. I had recently moved to Silicon Valley and didn't know that Ames had a supercomputer or I'd have gone to see it. So I missed it then, and was glad to at least get to see its remains on respectful display at a site that could appreciate it.
The Illiac-IV that went to Ames wasn't originally slated for NASA. It was built for the University of Illinois, but this was a time of student unrest, and while the machine was being built, a bomb was exploded at the University of Wisconsin at Madison. It was decided that it might be prudent to ask for proposals. It turned out that Ames, whose site perimeter was patrolled by U.S. Marines, had the most appealing proposal.
Ames definitely had use for the machine. A lot of wind-tunnel work was being done there, experimental studies of turbulence around airfoils. If there was ever a use for a massively parallel computer, this was it. The first serious program run on it did a simulation of the flow of air around the nose of the space shuttle. The machine was so unreliable then that they replicated the program, ran the replicants in parallel, and let them vote on the right answers. Much of the research done on the Illiac-IV was of this sort, highly parallel, although not always for redundancy. The NASA Illiac-IV proved for the first time that science could sometimes be done by simulation, that it wasn't always necessary to do the experiment. This was especially important for Ames, where "doing the experiment" often meant calling up the power company to negotiate for a time when Ames could draw 100 megawatts off the power grid to start up a wind tunnel.
I learned all this from current and ex-NASA Ames researchers, who spoke in a dusty warehouse space at Ames, right next door to the space given to the model railroad club. The warehouse space is what the Computer Museum is using until it gets more appropriate space to display its collection of old hardware.
I saw a Cray to walk into, the inside all blue spaghetti wires, an Osborne 1, a Radio Shack Model 100, the first laser printer from Xerox PARC, an early line printer. Under glass was a 1975 Adidas running shoe with a microprocessor and digital readout in its tongue so you could monitor your speed, distance, and caloric output while you ran. Next to it was a red, white, blue, and green Apple Powerbook 170, personally autographed by John Sculley. Sorry, John. Fame is fleeting. Adidas was hot once, too.
The Illiac had a megabyte of RAM, and, according to the engineers who worked on it, was the precursor to everything happening in parallel processing today. That sounds a little grandiose, especially since various models of parallel processing are being explored and exploited in microprocessor design and Illiac-IV was a pretty straightforward SIMD design.
It was also a legend. It was 65 feet by 12 feet and pulled 3/4 of a megawatt of power; less than a wind tunnel, but considerably more than even a fully-loaded PC. This particular machine was host 15 on the Arpanet. Almost everyone who used it used it remotely, via 300-baud acoustic couplers. One of the more famous names of people who used it is John Warnock. It was used for the analysis of satellite image data and did some of the computation for the Galileo Jupiter probe.
At one point an audience member asked, deadpan, "Can you comment on any of the top-secret operations?" As it turns out, although Illiac did help NASA researchers locate a submarine where no submarine should have been, the top-secret work was minimal, because it "took six hours to switch from red to black" (from secure to unsecure mode). So it wasn't used for a lot of top-secret stuff. Or so they said. When it was decommissioned, it was replaced by a Cray.
Finally, here's a simple little puzzle. I wonder if anyone will get it. What's the connection between Steve Jobs and the phrase "anywhere but here"? If you know, you can e-mail me at mswaine@ cruzio.com. As usual, no prizes will be awarded.
DDJ