Canon S90 Review

Cazedero LDT 2010

I ordered a Canon S90 for this weekend's trip, hoping that Lisa would be able to capture Myers Grade, one of the prettiest roads in the area, but one that's impossible to do justice to because descending on an 18% grade is challenging enough without thinking about taking pictures.

Well, that didn't work out, but the camera did! First, I was impressed by how much control the camera gave me. You really do get to set aperture, shutter speed, etc., and the manual focus option is even useful (the display zooms in so you can tell whether or not you're out of focus).

Others complained that the control ring at the back of the camera is loose and easy to turn by accident. In practice, that's not an issue for us. Neither Lisa nor I tripped it over the weekend. The RAW mode is very nice, but I've noticed that in full Auto mode (which Lisa uses), it doesn't shoot in RAW. Nevertheless, that's a minor issue. RAW files are huge (8GB only stores 540 RAWs), so for shooting off the bike (which is what Lisa does), I doubt if we'll use RAW: there are too many bad pictures to be thrown away!

In any case, the camera is great, so I expect that we'll keep it past the trial period and rely on it for the upcoming tour of the German Speaking alps.

Hoisted from my Buzz

This started off as a post on Buzz, but then I realized that most people probably neither read nor follow my Buzz/Reader feeds (which is admitted a torrent of information that more than one person has complained about not being able to follow).

Google has just handed out a Founders' Award to the Android team. By all accounts it was substantial, ranging from a pat on the back to millions of dollars in restricted stock (vested over 4 years, of course --- Google is not stupid). Yet Google has seen two recent departures (neither of which I know well, so this is all outside information) from the Android team. Presumably, if the stock was vesting and it was substantial, people wouldn't leave now, yet precisely that's what's been happening! Why is that?

One possibility is that the people who left didn't get to take part in the substantial stock awards. If Google handed out a million bucks to your cube-mate and not to you, that's a pretty strong signal that you should leave right now. This obviously applies to you if all you were was a Senior Software Engineer who has next to no prospects of promotion under the current regime. I find that hard to believe, however, of the Senior Product Manager. There are a lot fewer PMs than there are Engineers, so the chances of a low payout is less likely there (but it could still happen!).

The deeper possibility is that not all $1M awards are the same. A $1M award out of Google vested over four years is the same as a $250k/year increase in salary. But that's going to be taxed at your highest marginal tax rate, somewhere around 45% in California. To make things simple and to account for the possibility of a tax hike, I'll call that 50%. That's a $125k/year increase in salary. That's great, but let's take a look at the alternative if you were getting $1M from LinkedIn. (My sources say that LinkedIn is worth about $4b-$5b today)

If you've read my book, you'll do the right thing and immediately pre-exercise all your LinkedIn options. 4 years later (or whenever you can), you could sell the stock for $1M. However, that's all taxed at Long Term Capital Gains tax rates, which are currently at 15%. Tack on another 10% California taxes, and now you're at 25% tax rates. So that nets $750,000 over 4 years. (Note: tax rates can change) Compare this with $500,000 over 4 years, and the person who joined the startup gets $250,000 more. That's 30% more money!

But that's assuming that the startup's package is worth $1M as well. The startup options carry a lot more uncertainty, and LinkedIn is much more likely to double over the next four years than Google. It's also much more likely to go to $0 (No risk, no reward!). Furthermore, you can control when you extract income out of your LinkedIn stock (through an 83b election as described in the book), whereas Google's Restricted Stock gets taxed as you vest.

Of course, the best deal is to do what these gentlemen probably did, which is to use the founder's award to extract more equity from the respective companies they joined, as well as promotions, salary increases, and other benefits. On top of that, the bigger impact you make as a person in a smaller company also makes the potential career path there much more satisfying, provided you did your homework when you selected the startup. Note that you can also have a very satisfying career at Google by being on the fast track and getting promoted every year --- Andy Rubin wouldn't join Facebook any time soon. But if you're not one of those, then Beust's and Tseng's approach makes a lot of sense.

Review: The Prestige

I first read the novel The Prestige some time around 1998 or 1999 through a friend I met on the train. When I heard that Christopher Nolan had made The Prestige [Blu-ray]into a movie, I made up my mind that I would like to see it. But things being what they were, I didn't get around it until now.

Most movie adaptations of novels are lackluster. I wondered how the movie would do in this case, partly because the prose of the novel was turgid because the novel also aimed to pull one over the reader's eyes as well, while remaining fair enough an intelligent reader could figure out what the Prestige was. (The prestige is the final act of a magic show, where the magician produces whatever he made disappear) And yes, I did figure out what the prestige was, so it was fair.

Well, the movie does reproduce the novel remarkably well, and once again, an intelligent viewer who pays attention will definitely has all the clues he or she needs in order to figure out the trick. The actors are great, and I definitely loved the portrayal of the rivalry between Jackman and Bale's characters. There are a few changes from the novel, but none of them are outrageous.

I think the biggest problem deciding between the novel or the movie is that you can only experience the reveal once, in whatever form. Once you've read the novel, then the movie won't hold any mysteries for you, and vice-versa, so you have to be very careful in deciding which you should sample first. Given that the prose of the novel was deliberately turgid, in this case I recommend that you watch the movie first, which will likely motivate you to spend the extra time required to read the novel. If you think that means that I have high praise for the movie, you're right. I'm quickly becoming a Christopher Nolan fan!

Review: The Hunger Games

I started The Hunger Games this morning and finished it in a day. It was an incredibly compelling read for me, making me understand why my mom found reality TV so compelling.

The story is about a girl, Katniss, who volunteers to take the place of her beloved sister for The Hunger Games, an annual event at the Capital of her country (set in a future North America) where 24 children are brought together to fight to the death on TV in a Survivor-like environment. Suzanne Collins throws in twists like sponsors of the various contestants being able to send gifts to the participants in time of need, and the game-masters being able to control the environment to some extent.

The protagonist is what you would expect a heroine to be: resourceful, talented, strong, and smart. The action is fast-moving and quick: at no point is the reader even left time to be bored, contemplative, or able to question the premise. This could easily be turned into a movie, though perhaps one that's a bit too brutal to watch if depicted in all its glory. You really do feel like you're watching as the protagonist manipulates the environment, her opponents, and even her audience to survive challenge after challenge.

All in all, a good read, and I guess I'll be checking out the next book from the library.

Guide to Startups Presentation Gets Audio Presentation

Unfortunately, I did not record the talk at LOLapps last week, so I've recorded a separate talk to go with this. Unfortunately, the best part of talks is in the questions, so that's entirely missing. But if you do want to ask questions, you can always join the book's mailing list.



Download MP3

I hope the audio is listenable!

Review: The Windup Girl

The Windup Girl(DRM-free ebook) is Paolo Baciagalupi's novel about a post-fossil fuel economy and dystopia. I stumbled across it by reading two short stories set in the same world (free download).

Set in Thailand, the novel poses an energy poor future, one in which fossil fuels are expensive and almost non-existent, where elephants wind giant springs to store energy, and bio-plagues released by corporations have swept the world, wiping out any sources of food not genetically engineered to resist the plagues. Genetic engineering has become far advanced, enabling the creation of beings who are subservient to humans. These new breed of slaves are called windups, since they have jerky motions genetically engineered into them to make them distinguishable.

The plot itself is complicated, and seem designed more to show off the milieu than anything else. The characters themselves are not, with fairly simple motivations to drive them: many are even caricatures that you would expect.

What shines in the novel, however is the world-building and the verisimilitude. Most novels set in South-East Asia are terrible, poorly researched and with no regard for the culture and interactions between the mix of races. Liz William's Snake Agent was the most horrible offender, but most aren't much better. Bacigalupi, however, must either have really lived in South-East Asia, or has done such meticulous research that no one else has done before. An early MacGuffin, for instance, is easily identified by a South East Asian as the Rambutan, but few non-Asians have heard of the fruit, let alone think to use it as a MacGuffin.

Another sign of the incredible work Bacigalupi has done is to note the tension between the local Chinese and the native races. It's accurate, intelligent, and very believable. Even the names of the characters (one of them is called Tan Hock Seng, something you'll only find in South East Asian names) correctly indicate where they are from and what their likely cultural attitudes are. The language, when it switches to Mandarin or Hokkien, is also recognizable and correct. This novel was clearly written by someone who respects South East Asian culture well enough to get things right, and that in itself is highly commendable.

At the very least, I think you should read the short stories, and then decide for yourself whether or not you want more Bacigalupi. My big disappointment is that I don't think the novel is as great as the short stories are (the short stories aren't all set in Thailand, but do have elements also found in the novel). On the other hand, the exploration of the themes, and the accuracy of the novel with respect to its depiction of South-East Asia compels me to recommend it. Well worth your time, if you've got any interest at all in these topics.

[Update: The Windup Girl has won this year's Nebula Award for best novel]

Review: Garmin GSC10

After almost a month with the Garmin Edge 500, I noticed that even the high sensitivity GPS chip tended to have odd spikes. For instance, it claimed that we hit 55 mph on Skyline, when my same bike computer said we did not exceed 45mph. On my single, I would not care, since I'm hardly the kind of person who sets speed records anyway, but on the tandem, we've hit freeway speeds, so I found myself wanting some precision, especially for the upcoming Tour of the German Speaking Alps.

Now, even regular bike computers don't stay accurate because of tire inflation variation, air temperature, and just plain inaccuracies while calibrating the wheel size. However, the Garmin GSC 10 Speed/Cadence Bike Sensor does the right thing, which is to calibrate the wheel size using the GPS unit, and then use the spinning wheel as a check against GPS jitter. On top of this you get a cadence sensor as well, which produces nice charts telling me that keeping my cadence on the tandem up on Redwood Gulch would be a problem. [Update: Our Mt. Charlie Ride showed us doing 79.1mph. I don't find that believable at all, so there's still some jitter!]

Mounting and setting up the unit was a cinch. The unit cleverly has both the cadence and the speed magnet arm on the same sensor. The arm is designed in such a way that you can twist it up and down without needing a screw driver to loosen it up. This is a big feature, since if things get twisted around on a ride somewhere you can fix it without getting out the tool. Checking to make sure that you got everything lined up is straight forward: you push a reset button and the LED blinks every time a magnet crosses the sensor arm, so you can tell whether or not you've got it right. The unit uses CR2032 batteries Pack, which I had a stash of back when I had an operational Sigma MHR 2006. A search on the web indicates that the battery should last about 2500 miles, which sounds kind of low to me, but given that everything is wireless is perhaps understandable.

My biggest complaint is that mounting the unit uses zip-ties, instead of rubber bands or something less permanent. My experience is that zip-ties are prone to breaking off from fatigue, and of course getting snapped through abuse, and I'm unlikely to carry zip ties on tour. The battery cover is also in an awkward place and might require removing the rear wheel to replace. No big deal if you're not a cycle tourist, but still a pain.

All in all, it's a nice package, and I'm even tempted to get one for my single. I'll probably wait to see whether this thing survives the upcoming tour before I buy a second one for the single bike. Otherwise, first impressions are: recommended.
[Update: I just switched the battery, and it's been about 1500 mile per CR 2032]

Review: The Other Lands

The Other Lands (Acacia, Book 2) is David Anthony Durham's sequel to Acacia, which I reviewed 3 years ago.

Acacia ended with a war over, and Corinn, the most morally ambiguous of the Akaran children Queen. We see the aftermath of the war, and Corinn's rule resume everything that her other siblings considered wrong about the rule of Accacia: slavery.

The plot then broadens to include the rest of the world we had seen merely a glimpse of in the first book, the land of the Numreks. Books that in the middle of a trilogy tend to suffer, since the novel typically tries to set up the finale, and not much happens. While this book does end with a cliff-hanger that the long story arc sets up, you can't accuse Durham of making nothing happen in this book. Lots happen in this book, and we see many mysteries resolved. For instance, why was a seemingly endless supply of slaves needed? What were the Numreks running from?

My only possible complaint is that the prose while still very readable, isn't as sparkling as Acacia was. Nevertheless, I found it a good read, and the first fiction book I'd finished for quite some time. Recommended.

While doing research for this blog entry, I discovered that David Anthony Durham won the 2009 Campbell Award for best new writer of Science Fiction and Fantasy. Congratulations. And yes, Scarlet, I still think you're wrong about Acacia.

Promotion Systems Redux

My spies at Google tell me that my essay on promotion systems drew a reaction on a Google mailing list that I didn't read even back when I was employed at Google.

I did get a few interesting responses. I won't post them here, but I'll address what a few points: The first was that the system isn't actually the worst of all systems. An even worse system would have all the characteristics I pointed at, and add one more feature, which is to have promotions be cut-throat (i.e., for you to win, someone else would have to lose). The peer review feature does make people nicer to each other, though what I found as time went on was that if you were too low on the hierarchy, your opinions didn't matter to the promotion committees anyway, so more cynical senior people would feel free to ignore your requests for help.

Someone else pointed out that the real compensation at Google was much more dependent on raises, stock and bonuses, so the promotion system didn't matter. But that's not true at all! First of all, there was a level-based salary cap, and people did hit them. So at some point, you have to get promoted if you want your salary to keep going up. Bonuses were based on salaries, so those don't go up if your salary doesn't go up. Secondly, even if it was true that options and bonuses was all that mattered, then you just validated the claim that eliminating the engineering ladder wouldn't hurt productivity at all!

Finally, people seem to have an inordinate amount of faith in a process executed by humans who had limited amount of time to read and evaluate a person's work. When I sat on intern hiring committees, I would dig through change lists and discover all sorts of nasty things about an intern's projects (the intern was rarely to blame, but his/her mentor was definitely at fault) that would lead to a "no hire." Yet I was usually the only person who bothered to do that kind of digging. Most people on the committee just read the person's packet and took everything at face value because it was way too much work otherwise. I've sat in on promotion committees where the director moderating the discussion (remember, the managers weren't supposed to influence the process) would provide out-of-band commentary by murmuring about how good a candidate was. I also saw cases where someone working on an un-sexy project was denied promotion just because the rest of the committee hadn't heard of it, and had no understanding of what the work was. I've even seen sympathy promotions, so I have no illusions about whether or not the system was fair.

Now, a traditional management/promotion system might not do any better, but you bet the heck that at least the employee would be confronting his/her manager and at least demand clear guidance about what he had to do to get promoted. An employee denied promotion by a faceless committee can't do that. This was what was on Reed's mind when he said he still favored a traditional system executed well.

In any case, I think it's very healthy for Google to have an internal discussion about this. But do I expect the system to change? No. The super-star rule I referred to in that previous post would prevent that. I did have a discussion with a VP about this. He told me that when he first joined Google, he tried to change the promotion criteria to better formally recognize leadership, mentoring, and the importance of spreading knowledge (technical or otherwise) throughout the organization. The result? A bunch of very senior engineers (who had all benefited under the current regime, and were understandably worried about their career prospects under a different system) shouted him down. If a VP can't change the system, no amount of blogging from outside or inside the company will, so these will be my last words about the topic.

San Francisco Startup Tour

Guide to Silicon Valley Startups Talk MP3

I gave a talk today at LOLapps.com, which is in a lovely location in downtown San Francisco: this is clearly a startup that believes in planning ahead: the entire 7th floor was theirs, and they looked like they had enough room to double or triple while staying in the same space. Folks look excited and asked intelligent (and sometimes difficult) questions, which is great, because that's really what makes the talk (otherwise, you could just read the book)

Unfortunately, the talk wasn't recorded, so I'll just have to do it on a microphone at home one of these days and put up an audio to accompany those slides.

I didn't intend to, but afterwards I met with a friend who had just taken an offer from another startup, and then headed over to visit NextStop, which is a 6 person startup down near the Caltrain station. I was surprised by how empty their building was: the downturn seems to have decimated office space occupancy in San Francisco, but that just means it's a great time for startups, though I'm seeing signs now that the Bay Area is heating up again.

All in all, a busy day, with lots of ideas to ponder.

Engineer's Guide to Startups back in stock!

The second printing of An Engineer's Guide to Silicon Valley Startups is now available. I've also updated the book's web-page to point to some of the recent blog posts that some may find interesting. If you've been holding off because you wanted a paper copy, you can now place your orders.

Thanks again to everyone who's helped make this book a success!

Leadership Training

After I was done with the Edge workshop, one of the organizers asked me if I would be interested in follow-on workshops. Hoping that it would be more practical than the usual executive team building type off-site, I agreed and decided to go.

Well, it turned out to be less of the typical executive team building offsite. For one thing, it was entirely indoors, and for another, it was focused on achieving your goals. Now, I'm the kind of person for whom achieving my goals is actually easier than defining them, but I was curious as to what value the person had to add.

It turned out to be a peer-based setup. And the group was filled with Google old-timers, each of whom would stand up and talk about their goals, many of whom weren't at all related to business objectives. For instance, Patri Friedman talked about Sea Steading.

The saddest thing I ever heard at Google, however, came out of this workshop. One of the participants stepped forward and said, "I've now been at Google for 7 years now, and I have all the money I ever need to never work a day job again for the rest of my life. I would like to quit and work on *charity redacted* full time, but I'm too afraid to to do this. You see, I've leaned on having engineers around to help with my projects my entire career at Google, which was my first job out of school, and I don't know how to get things done without them."

I was first stunned by the honesty and humility behind this statement. Most businessmen, sales people, and non-engineers cannot admit that without the ability of engineers to create products and maintain them, any business simply will not exist. In the console universe, you ship products built to a fixed platform (though on PS3 games, you frequently get loads of patches as well), you might be able to dismantle the engineering team behind the product after shipping. But in the new Software As A Service model? No freaking way. No engineer, no product.

The saddest thing about this is that this person could not conceive that with all the resources at his/her disposal, (s)he might be able to just learn how to think like an engineer and apply that to the problem that (s)he would like to solve. The thought just filled me with sadness.

I recently read John T Reed's Self Publishing, and he pointed out that most of his books were just about applying the engineering mind-set to the subjects at hand:
It's like the astronomer who wows aborigines by accurately predicting an eclipse. It's easy to do if you have an almanac. But the aborigines not only have no almanac, they did not know there was such a thing as an almanac... By taking my engineering training to various lands of the blind, like real estate, I become king--or at least a successful how-to writer.

I think very few engineers realize how rare this mentality and capability is, and how valuable it is (by the way, Google does realize how rare it is, and does hand out million dollar bonuses for that rare ability). And I guess that's a good thing. Otherwise, corporations would have to pay real engineers as though they were worth as much as the "financial engineers" on wall street. After all, a war for talent should be prevented.

Review: Predictably Irrational

Predictably Irrational: The Hidden Forces That Shape Our Decisions is Dan Ariely's book about his adventures with behavioral economics. It's a compelling read, but has also woven itself so much into popular culture that you may already be familiar with most of it.

The book covers many human foibles, with a chapter devoted to each one. These are:

• Comparison Failures: We judge objects and people by comparison, rather than by absolute worth, so we get anchored to the comparables rather than to true value.
  
• Anchoring: We get locked to the first number we see, even if it's utterly random. That's why Steve Jobs hawks the iPad by saying, how much would you pay for the device? $1000? $800? How about $499. So the $499 price looks like a bargain in comparison to the $1000 price.
  
• Free almost overpoweringly over-rides everything else. Even if it is a worse value than a non-free product. The experiment here was great. He offered people a $8 coupon or a $20 coupon for $10. Even though the $20 coupon was a better deal, most people took the $8 coupon.
  
• Social norms versus market norms. Basically, when you do a favor for a friend, an offer to pay is an insult. That's why companies try to motivate employees by trying to lock employees into social norm, so you feel bad for saying no to working on weekends (after all, you wouldn't say no to visiting your mom on a weekend). The problem comes when employers make business decisions and then cause the employee to feel betrayed (your mom wouldn't lay you off!).
  
• Decisions made in the heat of the moment. It turns out that we under-estimate our lack of judgement when we are in emotionally heated situations. That's why major decisions are best made after sleeping on them, or better yet, if there's a time delay between when you feel the emotion and when you make the decision.
  
• Overvaluing what we own. This one is all over popular culture. Once you own something, you feel much worse about giving it up than if you had never tried it in the first place. That's why free trials work.
  
• Analysis Paralysis. I see this all the time, when people try to keep so many options open that they fail to see that they would have been better off committing to one in the first place. My favorite example in this book was of the woman who still kept trying to keep her old boyfriend even though she liked her new one better. She felt bad losing the option to her old boyfriend, even though it was irrational to do so.
  
• The Placebo Effect. This one is so well known I was surprised he bothered to spend so much time on it in the book.
  
• Cheating. It turns out that we cheat less if we are reminded of the usual high minded values before being exposed to temptation, and we cheat more if the cheating is one-step removed from tangibles. In other words, it's way easier to steal credit card numbers than it is to steal cash, and it's even easier to sell CDOs and bad mortgages than to actively bilk you of your money.
  

Quite a lot of topics were covered, but and the book's an easy read. The only disappointment, I guess, is that so much of it already permeates popular knowledge, so it's easy to go through the book nodding at every chapter and then asking, "Why does it take so many experiments to show what we already know?" Of course, that's precisely the state of behavioral economics: it's so new that smart people like Ariely have to prove what should be first principles. Ultimately, I recommend this book, but I do hope that in the future, we see less obvious irrationalities from Ariely's research. I'm glad I borrowed it from the library, because I would have felt about paying even the Kindle price for a bunch of stuff I already knew. If you don't have time to read the book, watch Ariely on YouTube:

Loyalty is a two way street

I was recently getting together with a friend/former colleague from a previous startup. We were discussing engineers who had spent a long time at the startup (which never went anywhere). It brought back to me Mark Suster's article about job-hoppers. It reminded me that loyalty has to be a two way street. Yet companies and employers lay-off long-time employees all the time as business decisions.

You can't have it both ways. You can't claim that the company is a family and that you owe it to a company to be loyal when someone else comes by with a 30% raise, and then turn around and layoff employees when business gets rough. If a family member did that to you, you'd feel incredibly betrayed, and the typical rank-and-file worker has been continually betrayed in this fashion for the last 30 years or so. An investor or venture capitalist can invest in 20 startups at once, and if 19 of them fail and the 20th turns out to be the next Google, the investors still win. An employee/entrepreneur can work in only one startup at a time. If after 3-4 years, there's no exit in sight, and you're not being granted new options, a hefty title, or some other compensation, then I would start looking for a new job! Note that many startups don't have an active retention program in the form of refresh options for employees (even high performers), and I think that's a major mistake. Obviously, a startup founder can't just leave, since if he leaves that's a signal of major disaster.

In retrospect, the time when every one of us employees should have exited was when we saw the Chief Architect leave. He had all the incentives to make the startup successful, but if he was leaving, then what reason did the rest of us had to stay? Incidentally, he eventually ended up at Google, where he got handsomely rewarded, so his departure was well-timed. I discuss other reasons to leave in An Engineer's Guide to Silicon Valley Startups. By the way, the print copy is current sold out, but digital editions are still available, and I expect the second printing to be available next week.

WM2016 Followup

Our LG WM2016 somehow became unleveled, so I called the GeekSquad and had them come over. The technician first showed me how to level the machine myself, and then told me about several issues:

• Don't use bleach for cleaning the tub. Use dishwasher fluid instead. That gets rid of oil, which bleach doesn't do. Run the machine in "Hot Wash"/Cotton mode. You only have to do it every 2-3 months.
  
• Hair tends to accumulate at the bottom of the glass door. You must clean off the hair, or it will eventually cause a door leak.
  
• When using detergent, use half of the "1" line on the Tide cup. The machine is a small machine, and using too much detergent would eventually destroy the machine. A 96 load version of the Tide bottle will last 1-2 years.
  

Wow, I didn't know quite so much about the machine even after reading the manual. I guess the extended warranty/GeekSquad coverage wasn't as much of a waste of money as I expected!

Diet and exercise don't work?

In the Atlantic, Marc Ambinder wrote:Stigma might be more bearable—an unpleasant way station on the path to a thinner, healthier life—if diet and exercise, the most prescribed solutions to obesity, worked. But they don’t. Qualification: if you eat less and exercise more, you’ll lose weight. But the chances that you’ll stick with that regimen are slim, and the chances that you’ll regain the weight, and then some, are quite high. A systematic review of weight-loss programs, by Thomas A. Wadden and Adam Gilden Tsai of the University of Pennsylvania, found that the evidence that commercial and self-help weight-loss programs work is “suboptimal.” People who diet often regain more weight than they lose.

I can think of two counter examples. A friend of mine at work was diagnosed with a heart problem. It was exacerbated by a sedentary lifestyle, and his doctor wanted to immediately put him on drugs. He said, "Wait, wait. Can I solve this with a lifestyle change? Change my diet and exercise?" His doctor replied, "That doesn't work. Statistically, nobody sticks with such regiments." My friend wasn't willing to give up, however, and told the doctor, "Let me try it for a month. If that doesn't work you can put me on drugs." This guy went from zero exercise to biking to work 4-5 days a week, hiking and running with his kids on weekends, and started cutting his portion sizes and eating more greens. A month later, his doctor pronounced him completely fit, and at little risk from his cardiovascular disease. Six months later, he was still going strong, and still biking to work nearly every day. Note that if this same man had lived outside California, he probably wouldn't be able to bike through the entire winter. If he had even lived in San Francisco, his bike probably would have gotten stolen within that time period.

Example two: In 2005, I was diagnosed with osteopenia. My doctor immediately put me on a regiment of calcium and vitamin D supplements, and I embarked on a program of hiking and weight lifting, that I continue to this day (it's been 5 years). My bones are back to normal, even though doctors and others were incredulous at the improvement.

The lesson, perhaps, is that if you're a Google engineer, all these rules of thumb about lifestyle changes not working? They're probably inapplicable to you. The question in my mind is: Is there an easy way to predict what kind of persons do well with this kind of diet/exercise regime, and what kinds of persons don't?

Review: Up in the Air

Ok, so this review is actually a bait and switch. Last night, Lisa wanted to watch a movie. I had neglected to sign up for Netflix, but then remembered that the PS3 had on-line movie rentals through the Playstation Store, so I fired up the game machine and looked through the store looking for something to watch.

It wasn't hard to find something to watch, since Up in the Air [Blu-ray] was a movie that everyone on our St. Vincent Trip talked about. It felt odd not to have the Amazon reviews available while browsing, but I got over that since we wanted to watch the movie.

Purchasing is a bit of a snag, since you can only buy Playstation Store dollars in units of $5 at a time, but I could see myself making regular use of this feature, so there wasn't a lot of resistance on my part. Once the download started, I could see that my WiFi connection while good, wasn't going to instantaneously download the movie for me to watch right away. However, once the PS3 had downloaded about 5% of the movie, an option appeared to watch the movie while downloading. It took no more than 3 minutes to get to this point, but getting Lisa ready to watch the movie took at least 10 minutes, so it was no big deal.

The nice thing about watching the movie this way, as opposed to say, Redbox is that there are no ads. You click play, and the movie immediately starts. No trailers, no menu, nothing. Just what all online rental experiences should be like. The picture quality is absolutely amazing. As mentioned before, the PS3 upscaling works smoothly and magically, and Up in the Air wasn't the kind of movie that needed HD anyway.

Twice, however, the playback stopped when the playback went faster than the download could handle. Each time, we'd stop and do other stuff for a bit so that we wouldn't get stuttering. I think if I had about 30 minutes of buffer time before watching the movie, I don't think we would have even gotten a pause. I would certainly do this kind of rental again, as opposed to renting from Redbox or paying a monthly fee to Netflix (though as a cheapskate, I would of course get a free trial from Netflix first). For HD movies, however, there's some planning needed since it would take quite a while to download the entire movie to disk. That's probably better done through a regular Blu-Ray rental or Netflix.

As for the movie itself: I liked it. Unlike Hollywood movies, there's a plot twist that doesn't end happily for the protagonist, which I thought was unusual. Worth the rental, but I don't think we needed to see it in the theater.

Guide to Startups Visits San Francisco

LOLapps, a startup in San Francisco, has invited me to give a talk there on May 5th. It will be on-site, and the talk will be open to all comers, including non-employees and walk-ins. (I'll post contact info for the talk later) I certainly didn't expect to give a talk at a San Francisco startup before one in Silicon Valley, but "Silicon Valley" has really come to include the greater Bay Area in recent years. The Berkeley area, for instance, has always had a fair number of startups that became successful, including Inktomi, Ingres, Sybase, and Perforce (Perforce was mentioned in my book).

I'm expecting to keep the talk short (30 minutes or less) so as to have time for questions. If you're in the neighborhood and read the book (or haven't read the book, but want to find out more), please show up and ask questions.

The location is:
116 New Montgomery
Suite 700
San Francisco, CA 94105

The cross street is Mission, and it's 1 block away from Montgomery Street station

I'm down to a little more than a handful of print copies of An Engineer's Guide to Silicon Valley Startups left. I've placed an order for a second printing, but it usually takes a week or so for it to show up. Thanks to all of you for helping make this book a success!

Review: Perfect Rigor

Perfect Rigor is Masha Gessen's book about Grisha Perelman, the mathematician who proved the Poincaré Conjecture. It's a short book and a compelling read: I got it home from the library at 5pm yesterday, and finished it at 11pm, with a couple of hours off for cooking, eating, and mowing the lawn.

Perelman doesn't do interviews, and hates talking to people, so Gessen was forced to go about writing this book by interviewing his friends, teachers, and associates. Perelman's childhood was filled with mathematics, having been enrolled by his math tutor into a math club, where Perelman did very well. The book does a great job at describing what it was like growing up in the Soviet system for mathematicians during the 80s, with widespread anti-Semetic sentiments. Perelman was brilliant, however, and was lucky in having great mentors who nurtured and developed him as a problem solver. This first part of the book makes for great reading.

After Perelman leaves the Soviet Union, however, the book flounders a bit. First of all, Gessen had to describe that Poincaré Conjecture to laymen. This is not easy to do with text only: there are no diagrams throughout the book. I feel that this is the weakest part of the book, and Gessen made a hash out of it. Fortunately, the Wikipedia entry is thorough and does a good job of explanation.

Because Gessen did not have access to Perelman, we never understand Perelman's view of the whole thing. Why did he decline the Clark medal? We are led to believe that Perelman felt that he did not get the recognition that he deserved. Gessen also implies that Perelman might have Asperger syndrome, or was so idealistic that when the politics (again, really sociology) inside Mathematics came into play with regards to his solution of the problem, he decided to resign from Mathematics rather than put up with the problem. Perelman also declined the Millenium Prize.

Given that even his former mentors and teachers don't really have access to Perelman, we'll never know the complete story behind this.

20 years ago, when Fermat's Last Theorem was declared proven, I remember sitting down at dinner with one of my CS instructors. He told me about the result, and I said, "OK, so does this prove P=NP or anything interesting like that?" "No," came the reply. "Well then, I don't see why it's news." "You have no soul, Piaw!"

Unfortunately, I still feel like that. I don't think Gessen explains why the Poincaré Conjecture was important or interesting, and unfortunately neither this book nor Wikipedia does a good job of explaining the motivation behind the drive to solve this problem, which even after Perelman's publication, took teams of mathematicians nearly two years to fully explain. Nevertheless, a fun read, and worth the 3 hours of your time to read for a good understanding of the sociology behind Mathematics. Not that I could perform at the level these guys regularly do.

Review: YOU: The Owner's Manual

YOU: The Owner's Manual, Updated and Expanded Edition: An Insider's Guide to the Body that Will Make You Healthier and Younger was on sale in the Kindle store for $2, and I read the excerpt and liked it, so I bought it and devoured it over the course of a week. It's miles better than the previous health book I read.

The book considers the various pieces in your body that are vital to on-going health. Much likethe Nintendo game Brain Age, the book takes the view that you have a RealAge, which may or may not correspond to your calendar age. Everyone knows that a car that receives routine maintenance and oil changes will last much longer than a car that's been abused and not been maintained.

A lot of emphasis is on things that you can do. For instance, the authors explain why your stool is brown, and then tell you what's going on if your stool is of a different color, as well as what happens if there's blood in your stool. There's even an FAQ at the end of the book, covering questions readers have sent in for the earlier edition of the book.

Now, nearly every health book covers diet and exercise, but this book really does the right things. For instance, it recommends 3 sessions of 10 minutes of weight lifting a week, since that's apparently enough to provide good health. The book discourages you from running a marathon, since that's bad for your joints. In fact, exercising too much is considered bad as well, because of the increased in oxidation. This is in fact documented, but of course, some of us like traveling too much to give it up.

The book also covers diet, and even goes as far as to provide a comprehensive diet plan, complete with recipes, how to choose how much to eat, right down to how you snack and when you snack. This is very impressive: the entire book is a one-stop shop. If you can't be bothered to read any other book on health, this will do you very nicely. All the latest research is in this book as well, including the studies on aerobic exercise improving intelligence.

The only criticism I have on this book is that sometimes, the writing tries to be too clever by half. The analogies and references to popular culture, celebrities, and so forth feel a bit forced, as though the geek doctor in them was trying to be cool. Similarly, the chapters all start with a series of myths that the doctors then debunk. The problem is, by the end of the chapter, most people will remember the myths, rather than remembering the debunking.

Nevertheless, these criticisms are minor compared to how good the book is. Highly recommended. This is a book actually worth owning and referring to, rather than just checking out from the library.