My Steve Story

Everyone at Apple has a “Steve Story”. These are usually harrowing tales of running into Steve Jobs on the Apple campus and having a conversation with him. The reason these are usually harrowing is not because people idolize “Uncle Steve” and get intimidated by his presence, but rather because Steve used to have a penchant for firing people seemingly at random (he still might, but there’s every indication to believe that things have settled down a bit). However, it’s rumoured that at one time someone followed him around to reassure people that they weren’t actually fired and to return to their offices and keep up their good work.

I worked at Apple as an intern during Summer of 2007, and I had a royal blast. One of my projects shipped with Java6 on OS X very recently, and I discovered last weekend that one of the students in a class I’m tutoring actually used it as part of his project when it runs on OS X.

The feature, for the curious, was the JSR223 AppleScript support through javax.script, the scripting language support library only in Java 6. I originally started off by playing with jasconn, but the implementation that I eventually wrote is completely different, without requiring any System.execs (it’s a JNI, so interfacing with OSA was done in C). Interestingly, the bug has more information than the release notes. Anyway, it’s super-cool and if you’re a Java developer working on OS X you should check it out. Extra thanks to my mentor and my surrogate mentor for their help, since when I left it wasn’t really ready to ship yet.

Getting back on track, this was around the time when Leopard was converging and the iPhone was getting launched. It was a very busy time.

The atrium of IL1 usually has banners hanging off the ceiling, one facing the front door, the other facing the interior of the atrium. When I started, the front banner was a simple black advertisement for Leopard, and the rear was a colourful, 60’s style tie-dye iPod poster. At 6pm when the iPhone launched, the Leopard banner was taken down and a new iPhone one put up, also in a minimalist black style.

And I wondered: Leopard is going to ship before the next round of iPods, so why did the iPod banner stay up (instead of being replaced by the Leopard banner)?

So a week later, I had my chance to find out: I ran into Steve Jobs while standing in line for food in Caffé Macs with my surrogate mentor:

me: “May I ask you a question?”
Steve: “I’m at lunch”
me: “We’re both in line”
Steve: “… Okay”
me: “Why is the Leopard banner not up in the atrium of IL1 anymore? iPods are further off from being shipped than the OS.”
Steve: “(pause) Because the iPod one is prettier”

I paused for a moment, my brain parsing what I’d just heard, when I realized that Steve is not an engineer. He’s possibly the furthest thing from an engineer, which is exactly what Apple requires and needs from his position. The leader of the company isn’t necessarily a problem solver, but a problem finder. Their job: find problems the company can solve, and then sell the solution. When he talked to the entire company around the time the iPhone launched, he said that the motivation behind making the iPhone was because “everyone we talked to hated their phones.” It’s true, it seems so obvious when you think about it. No phone was really simple, easy to use, and powerful. Some came close, but every one had issues. But everyone hating their phones isn’t a good argument for Apple building a solution. Apple was the right company to create the solution when he showed us how the phone has very few unknowns in terms of Apple’s product experience because of our work with Macs and iPods already. And when you think of the technology in the iPhone, he was right. There’s an incredible amount of overlap.

All that thinking took about a second, at which point I thanked him, collected my food, and walked back to my team, who were all very surprised that I still had a job.

I think Steve’s a fine guy. Nothing to be afraid of, just a bit misunderstood. Apple’s going to be in serious trouble when he’s gone, but I plan on enjoying the time that he’s with us, because it’s going to be awesome.

The Uncelebrated Hacker

As per my previous post about losing living history, and my addiction to stories, I have another quick lament.

Recall the article on Richard Feynman from the previous post. This was a story about a Nobel Laureate written by a pretty successful guy who worked with him. Big stuff there.

Before I get myself into trouble, I don’t mean to say that Feynman wasn’t incredibly smart, he was one of the smartest men the world will ever see. However, there are other boffins that go uncelebrated, and their stories are no less important than those of the Nobel Laureates and people who have algorithms and numbers named after them.

As Computer Science has become more mature, I think that there are a more crazy hackers that go uncelebrated, even unknown, today than ever before. This happens in every industry as it develops. There are a few reasons for this:

• Engineers in our field have shifted away from solving new problems, to implementing solutions to known problems in new places because most of the problems we come across these days are thoroughly solved, which is much less interesting. Just as most work in Physics is arguably implementation of existing theories and not coming up with new ones, the computer field is no longer the Wild West frontier it used to be. There will still be neat developments (I think ZFS is one of them), but there won’t be as many, even though there are at least as many brilliant people in the field.
• There are also more engineers now than ever before, so standing out in the crowd is much harder. The smartest people don’t tend to publicize themselves (they prefer to use their time to be awesome instead), and it’s hard to take the time to recall what happened when you are around someone who mentally outclasses you, so when you retell a story it usually sounds like “wow this guy is smart” instead of “here is how/why this person is smart,” the latter of which is more interesting. This is probably why most of the best stories are either told firsthand in speech, or secondhand in writing.
• Another part of the problem is that it’s harder to get an algorithm or number named after you anymore than it used to be (an artifact of the computer field maturing). Even when there is an issue to be solved, the problems we run into now usually don’t have very simple solutions. Compare the implementations of ZFS and the Fast File System (UFS1). FFS was theoretically interesting, but implementing it was not very difficult. ZFS was also theoretically interesting, but the implementation required a team of engineers.
• Nowadays, you have to be more than just smart to invent something in computers (or any mature field), you have to be able to find problems to solve, and also be able to realize which problems are irrelevant or invalid.

There are a lot of things that contribute to the problem of uncelebrated hackers. There’s a lot to learn from their stories, if only we could hear them.

Who are the smartest people you know? What have they done? Any good stories?

Tell me a Story

I love stories. My favourite part of visiting old friends, my parents, or my family in Canada is not the free food, the hospitality, getting out of dish duty, or coming to a home with a made bed and the porch light on when I arrive bleary-eyed at 5:45 in the morning, it’s the stories. An afternoon working on the car, sitting on the porch with a beer, spending ‘quality time’ always results in me hearing a few stories about crazy antics I missed.

The only thing I like more than hearing about stories is making them, but in some ways that’s a lot more predictable. If I’m following the schedule of a normal weekday, it’s not likely that I’ll have anything great from that day to tell my kids about it in a decade or so, barring natural disaster or something crazy like that. It’s more likely that I’ll forget about the day completely. Let’s face it, if I go to work every day, I can’t have a great story from each and every one no matter how great my job is, and the rent won’t pay itself. On the other hand, when I go to a triathlon with my dad and our racing buddies, or to Canada to get in trouble with my relations, I always come back with a story. Since I can’t go out and make stories all the time, I’m addicted to reading them, maintaining a satiating flow.

The reason I like stories so much probably has its root in my interest in human development. All we are is the sum of our experiences, and stories represent the experiences that have had the most impact in shaping us, since we feel they’re important enough to re-tell. My parents and most of my family in Canada are people I’m very interested in. I would by lying if I called them stupid or dull, and the fact that they have qualities that I admire makes me very interested in finding out how they came to be that way.

Likewise, a few years ago I happened upon folklore.org, which is a repository of old stories from Apple during the “wild west” days of computer science, when any team of crazies in a garage could make a big difference just by doing what they enjoyed (in fact, that’s exactly what most of them did). The stories told on folklore.org are both brilliant and astonishing, and as a whole, the guys that worked on the Apple ][ series and the original Mac are really interesting to me. When I finally finished reading every single article, I still wanted more. I wanted to read about the Woz’ design decisions when he created the Apple I and ][, did Burrel keep his moustache, and what else happened at Apple during those years that wasn’t recorded?

Even more strongly, a year or two ago I read an article about Richard Feynman (which Daring Fireball reposted to my aggregator today, prompting this post). Reading it for the dozenth time didn’t take away any of its impact, and at the end I felt just a little bit hopeless, because Feynman is dead now, and while one of the last quotes by him in the article is, “when you get as old as I am, you start to realize that you’ve told most of the good stuff you know to other people anyway.” That’s really depressing to me, because I never got to hear all the “good stuff” he had to say, and I’m sure it was really interesting. This is living history, and we’re losing portions of it all the time.

I know I can’t do much about the loss of other people’s history, and I’m sure that my own stories and the stories I’ve heard will never measure up to the old stuff that I can’t get enough of, but I’ve decided to write an occasional storytime entry on my blog here to document my own history and interactions with people that interest me. Some will be (very) long, but all of them struck a chord with me, and I hope that they are also interesting to you.

Oh Hell No…

Sorry, Kia. You of all people are not allowed to name your car the soul.

Coyote on Security Theatre

I’m always confused as to why people care if you take a picture of the outside of a building or really at all. If I were a expert at espionage I certainly wouldn’t be holding the damn camera in plain sight. And beside what good is a picture of the barbed wired fence going to do? If you catch me rolling up in a van with infrared, thermography, and sonar telemetry equipment then you got a problem.

Why do authorities care so much when you take pictures of buildings? You’re not really getting any more useful information than you could with Google Earth or similar if you’re planning a large scale event, and you’re not getting very valuable information for something smaller, like intrusion.

Ever consider the possibility that maybe the building’s just pretty?

Microsoft Owning Yahoo: Only Good for Microsoft

For some background, check out this article detailing outcomes and options for Yahoo in the face of the Microsoft’s offer, and the possibilities for Microsoft to go hostile.

From the article:

We are learning that hostile takeovers have arrived in our industry. This is the second major hostile takeover so far — the other was Oracle’s takeover of Peoplesoft — but there will be more.

This is significant because historically hostile takeovers practically never happened in technology. Potential hostile acquirors assumed that hostile takeovers wouldn’t work because the target company’s employees would bail and the target company’s business would collapse.

It turns out that as technology companies become larger and more mature, acquirors are becoming increasingly convinced that neither of these assumptions hold. Perhaps employees of large tech companies aren’t that bonded to current management, and perhaps many of them would actually prefer to work for a larger, more dominant combined company. And maybe as a consequence, the target’s business would do just fine in the wake of a hostile takeover — in fact, maybe it would do better, due to advantages of combined size and scale.

My bet is that hostile takeovers, particularly of larger and more mature companies, are going to become increasingly common in our industry.

I think this is correct, but for the wrong reasons.

Mature companies aren’t necessarily valuable because of their employees and business. In the case of PeopleSoft (the first notable openly hostile takeover in our industry), many of engineers, the entire management, and many of PeopleSoft’s biggest customers left Oracle after the takeover.
In the case of Yahoo, I think the same sort of thing can be expected to happen. But Yahoo’s value to Microsoft isn’t people nor customers. Microsoft could buy Yahoo and fire everyone immediately, because

The value that Yahoo has to Microsoft is email.

Yahoo owns more than half of the world’s webmail. Microsoft is second with about a quarter, and GMail holds about 6%. Both Yahoo and GMail allow you to use POP3 or even IMAP (both open standards that work with any email client) to connect to your webmail account, which is handy. With Microsoft’s webmail, you can use a proprietary protocol that works only with Outlook (or Outlook Express) which work only with… Windows. While there is an Outlook Express (and an Outlook workalike, Entourage) for OS X, they don’t follow the GUI rules and conventions for Macs and are slow and difficult to use. If you’re on Linux, it’s possible, but requires odd hackery (this odd hackery works on OS X as well though).

So Microsoft is second biggest in the webmail world, and they don’t support any open standards.
That’s pretty expected. Historically, they’re either blatantly hostile to all standards, or they force their way through the processes (which really makes the standards process laughable, if you look at the saga of the ISO OOXML approval).

It’s more than webmail, though. Consider ‘real’ mail as well now. Exchange owns 62% of the corporate email environment (which, I’m sad to say, seems low). Exchange favours Entourage/Outlook for its clients, which use a proprietary protocol to make use of all the features Exchange provides.

The takeover of Yahoo by Microsoft gives Redmond a majority share in the world’s email. In effect, it gives them another monopoly.

The difficulty with resisting the hostile takeover is the pervasive myopic financial decisions made by public companies, and the courts arbitrating things like mergers, and the effect this has had on the law. From the article:

Suppose another bidder, like News Corp., enters the fray and offers $32/share, versus the current $29.68 Microsoft bid. Yahoo’s board would of course be free to take the higher bid from News Corp.

On the other hand, suppose Microsoft then raises its bid to $33/share, and then News Corp. holds its bid at $32/share. Could Yahoo’s board still take News Corp.’s bid in preference to Microsoft’s? In a word: no. When a board is presented with multiple offers, it can either take the highest objective offer or it can turn down all the offers. It cannot take an offer lower than the highest objective offer.

This is a phenomenally bad philosophy (actually, this would be enforced by Delaware law in the case of Yahoo). It completely disregards the (highly likely) possibility that Yahoo would improve its services, grow more effectively, and provide more value to its new parent under News Corp. than under Microsoft.

The takeover of Yahoo by Microsoft shouldn’t be stopped simply because it’s hostile and Yahoo doesn’t want to get owned by Redmond. While I believe this should be a good enough reason, it isn’t legally valid, especially since Yahoo’s stock doesn’t have a dual-class share structure that would allow them to repel boarders.

The takeover of Yahoo by Microsoft should be stopped (by the FTC) because it makes Microsoft a bigger monopoly (its status as a monopoly hasn’t changed since the antitrust suits against it were ruled, now it would just monopolize more), and should be stopped (by the shareholders) because it is bad for the company, and bad for their investments (unless their intention is to bail out and never look back).