Auto Ads by Adsense

Booking.com

Monday, September 30, 2019

Review: Training + Racing with a Power Meter

Now that I had a power meter, I decided to do some reading to see what it's supposed to do for me, and why it's considered "the high grade hotness" in cycling training. Training + Racing with a Power Meter came recommended (and is sold by) the people who made the PowerPod, but I was too cheap to buy the book and just checked it out from the library.

The book is a huge thick volume, and is chock full of diagrams, graphs, etc. It's mostly targeted for the coach, but obviously that applies to you as well if you're self-coaching. The graphs and technical what-not look intimidating, but I think I can distill the use of power meters in training as follows:

  1. You can use it to establish a baseline power called the FTP --- Functional Threshold Power, which is basically the highest sustained power you can maintain over a period of about an hour. The protocol for doing this is complex and sounds painful, but fortunately you can download an app to your Garmin watch/head unit that'll give you a fairly accurate assessment in a few minutes. Obviously, you can't do this without a power meter.
  2. In conjunction with a heart rate monitor, the power meter can tell you whether you're over-stressed (overtrained) or under-stressed (i.e., could stand to work out harder). Basically, if you can maintain or even increase your power output at the same heart rate as before, you can up the intensity/volume of your training without risking being over-trained. Conversely, if your heart rate is high but you have trouble sustaining your FTP, then you're over-trained and should back-off. This is actually of limited use to a self-coached cyclist, as you can judge for yourself whether you're tired. But hey, the data can be there to for you to recognized that you truly are tired, not that you're being lazy, and so might encourage you to rest more than you would otherwise. (Resting is apparently something very hard for the type-A race winning cyclist to do)
  3. Over time, you can see whether your power is increasing, which tells you how effective your training program is. Conversely, by checking your power output sustainability (basically, graphing your power output over time), you can find out how fresh (rested) you are. For instance, after a tour of the alps you might be very fit, but you certainly are not fresh. A week of rest later, however, your fitness might have dropped, but your freshness means you can go out cycling and break all your PRs on the local hills. Being able to track all that means that you can personalize exactly how much rest you need in order to have maximum freshness for a big event. (Known as the "taper")
  4. Finally, during the event, you use the power meter to pace yourself, so as to not exceed your FTP too often or for too long. Doing so burns one of your "matches" (and yes, you can use the power meter and a training regiment to figure out how many matches you have, and increase the number of matches you have), and you will not be able to sustain that pace for very long without damaging your overall power output.
And indeed, if you look at those 4 items above, you'll see that there's no reason to buy an expensive power meter --- the cheap unit (yes, it's in the book) will work just fine!

OK, did I just give you the short summary of an $18 book for free? Of course not. The devils are in the details, and the book mostly provides graphs, charts, and case studies on how to build a training regiment that's effective. It even classifies cyclists according to their power curve to be able to more easily customize an effective program for you and the type of riding you do. I discovered, for instance, that I was a sprinter-type (not a surprise: 23 and me told me that several years ago), but unfortunately, there's no training program that'll help a dad who has to carry 2 kids on a triplet over a 3 week tour. All the training advice seems geared to cyclists who are targeting a one day event, and who's racing for prizes, positions/whatever. I'm just trying to enjoy myself and have a good time while grabbing photos of the scenery, and there's no training program for that!

You don't even have to read the book to make use of the information. There are two major websites, Training Peaks and Xerts that will suck in your data from Garmin or Strava, and then either spit out a training plan to sync (Xerts) to your Garmin Edge automatically (watches not supported or I'd be tempted to try a 'structured workout') or let you buy a training plan/link to a coach (Training Peaks) to help you achieve your goals. Xerts claims to be powered by machine learning, and I wish they supported my Fenix 5X so I could see if their stuff worked, while Training Peaks seems to basically be an advertising platform for remote sports coaches to charge you hundreds of dollars a month to torture you, which is something I would think that I would pay to avoid, but of course, as one of my friends once said to me, "The problem with you, Piaw, is that you want to enjoy the ride! Don't you know that if you don't feel like throwing up you didn't go hard enough?" If that last statement describes you, don't walk, run to your web browser, buy the book, power meter, and head unit and go to it! If you don't think that enjoying the ride is a problem, then this blog post probably describes everything you needed to know about power meters.

P.S. It turns out there's a chrome extension (Elevate for Strava) that graphs and tracks your power development, fatigue, fitness, etc for free by scraping your data off Strava and storing it in a Chrome sandbox (i.e., it's not online, doesn't depend on servers and you have to manage saved files manually and back it up manually). One of my friends used it with great success for this year's PBP. It looks much easier to use but won't suggest workouts, etc to optimize your training.

Friday, September 27, 2019

Review: Shadow Captain

Shadow Captain is Alastair Reynold's middle book in the trilogy that started with Revenger. Revenger was very much a pirate story set in space, and Shadow Captain has the unenviable job of fulfilling both the pirate story part, as well as changing it from a typical space opera into more science fiction. It succeeds in the former, but fails in the latter, as the author decided to hold back more exposition in favor of having a 3rd book in a trilogy.

The start of the story is slow, though it makes sense. We get a nice picture of the Ness sisters spiraling into the life they'd hoped to leave behind, as they're pursued by the law and have to make questionable decisions. They each have been corrupted by the events in the previous book, and have to reconcile their decisions with their image of who they are.

The book picks up in the last 3rd, and there's plenty of action and the conceits of the science fiction setting finally kicks in --- we get the impression that the cycle of civilizations has a longer arc and is more malleable than we thought as the book goes on. I'll pick up the next book in the series: even mediocre Reynolds is still more interesting than most other people's fiction. Recommended.

Thursday, September 26, 2019

Review: Velocomp Powerpod V3

I'd resisted getting a power meter for years. To begin with, they're expensive, usually exceeding $600 when all is said and done. Secondly, most of them require you to either replace your wheel, cranks, or pedals. The pedal ones are particularly bad, since none of the power meters that are pedal based have walkable cleats!

When the Velocomp Powerpod v3 had a flash sale for $200, I decided it was worth a try. It was relatively light, was reasonably priced, and most of all, had technology that appealed to the software engineer in me. The approach Velocomp takes with power meters is to have a device with an integrated wind sensor, barometer, and several accelerometers that would also pair with your wheel sensor and cadence sensor.

The wind sensor would be used to calculate the opposing force via wind resistance. The barometer would be used to calculate opposing force from climbing (along with distance measurements from the wheel sensor). You would calculate rolling resistance from the deceleration when coasting (you would know when the cycling was coasting via the cadence sensor). You could even take into account a gravel road or bumps via an internal accelerometer that has a model for gravel. Tie everything together and you'd have a power meter that didn't need strain gauges. You could also tie everything together by automatically recalibrating at the start of every ride just in case the cyclist changed wheels, tire pressure, or even hand positions.

Of course, that meant that you couldn't have anything in front of the Powerpod that blocked the wind. Since I ride with a handlebar bag, the typical handlebar setup wouldn't work, but I'd already solved that problem for my light, so now I solved it by getting another Origin8 stub and a shim. The result was a pod that hung on the fork above the hub, in pretty much an ideal position.

The device has the mark of a home-made job. For instance, there's just one LED and one button. You pair your sensors by holding down that one button to turn it on. The LED is supposed to tell yellow so you can pair it with the wheel sensor (the cadence sensor is optional, but as indicated above, is also used to provide more data for calibration). But mine never turned yellow, no matter what. You can also pair a HRM sensor to it so that you can graph the HR data alongside the power. Then you're going to pair your head unit and go out for a calibration ride which is about 10 minutes long. You ride out and watch your power meter go from 0-50W, then turn around and ride back. When you get back to your start it'll read 100W and then start reporting real power. That's it! I recently tried calibrating it without following the out and back instructions, and to my surprise it calibrated fine as well. I guess it just needs a lot more data as the calibration required more time.

There are additional inputs you can provide for more accuracy, including your body weight and the weight of your bike. However, you don't have to stress very much about being absolutely accurate. This is where I wished the device would just sync with my Garmin connect account to extract body weight, for instance.

The device is extremely dependent on software, so software bugs would cause the power meter to lock up. At first I thought I had a defective unit or had broken it, but customer support told me to just hold down the power button for 15s to reset it. The reset didn't cause it to lose its calibration, but it did reset the internal clock so I had to tell it what the time was the next time I connected it to a PC (fortunately, the PC would also let me correct the date/times in the ride fall as well). In theory, the unit could be removed from the bike and then replaced, but every time you did that it'd have to recalibrate to its new position, so I simply solved that issue by bringing the laptop over and connecting it to download data. I suspect that the "bike-to-bike" move feature isn't going to be used very often, but so far recalibration hasn't been a major hassle.

Customer support answered relatively quickly, even on a Saturday or Sunday. So whatever else I can say, the company seemed to be very responsive. I thought at first that meant it was a one-man shop, but a visit to LinkedIn revealed that the company had about 15 employees or so, and had been around for quite a while. On reflection, the product itself contributes to the feeling that Velocomp is a one man shop. For instance, there's only a single LED, and it's blink states are used to indicate whether it's on, acquiring a pairing with a sensor, is low on power, or is charging. That's way too overloaded, and as a result I ran out of power once without realizing it, and I could never figure out when the device has been charged. I just end up plugging it in and leaving it plugged in overnight.

The device comes with software mysteriously called "Issac" after "Issac Newton", back when the product was called the Newton. The software graphs your power alongside your cadence, and can even merge your GPS files so you can visualize your power graph on Google Earth.

The big question people have is accuracy. I'm not about to buy a $600 power meter so I can verify the accuracy of the unit, but as far as I can tell, it's self-consistent. Obviously if you make a mistake by entering too high a body weight you'll get high readings for power, and of course, if you do an entire ride standing up but you calibrated the unit by being in the drops you might get power readings too low. But all in all, the unit has been very responsive and pretty good about adjusting for wheel swaps, etc.

Another issue with a device like this: what do you do with all this data? The big one as far as I can tell, is to integrate power into your training with a website such as Xert. These units will tell you (after you sync the data with Garmin Connect or Strava) what your fitness status is, how much power you have available, and even more intriguingly, how tired you are. You can then use that data to decide how hard to train and how much to rest. I was very surprised to see that even though I've always considered myself too lazy to seriously train, that I had a high enough volume of riding and physical activity that Xert and other that I was tired and should consider backing off or riding at lower intensity.

Immediately after acquiring the unit, I made my fastest time ever up a benchmark hill. I had the Garmin set to display power information, but went too hard at first and had to slow down. The power meter showed me sustaining a 200+W power output but as I got tired my power curve would drop. So if you're trying to pace yourself you'd figure out your long term sustainable power and try to stick to that instead of having power drop off, but it did show me how someone like a triathlete would use this data in a race. I'm likely to just ignore it and have fun.

The big limitation of a unit like this (as opposed to the pedal based or crank based system) is that it's only going to measure overall power use. So if you were on a triplet with your two kids, it could only measure the overall power produced by all of you, whereas the crank based unit would tell you how hard you were working. On the other hand, the wheel-based power meters also have the same issue, and you don't see anyone dismissing the legitimacy of those devices as power meters. You just have to be aware of the limitations of the unit before you buy.

All in all, the device punches well above its price class. I think if you want a power meter, it's well worth considering a powerpod over the more expensive alternatives. Certainly for training and racing purposes, most of the time you're not after absolute power, but relative power to see whether you're improving or need to back off. In which case, I see no reason to spend the extra money for a "real" power meter --- if you're serious about getting faster (I'm not), spend the difference on a coach or training service that will use the additional data to help you get stronger faster!

Wednesday, September 25, 2019

Review: Permafrost

Permafrost is Alastair Reynold's short time travel novel. It was expanded from one of his short stories, and is set entirely in Russia. Rather than deal with tricks or some sort of "law of unbreakable history" like some other time travel stories, Permafrost pushes head-on into time paradoxes, with interesting rules about what happens when a paradox is induced. The time travel mechanism is also interesting and indirect enough to be plausible, while introducing challenges not present with the direct time travel used so frequently in other science fiction.

The novel is too short to do much about character development, so the characters are basically cardboard cut-outs and aren't fleshed out. But the treatment of the material is fresh enough and the setting interesting enough that it was well worth the read. After all, even average Reynolds is miles better than most other writers' best work. Recommended.

Tuesday, September 24, 2019

Review: Revenant Gun

Revenant Gun is the last novel in Yoon Ha Lee's Calendaric Empire trilogy. In this last novel, we get an explanation of who the driving force is behind the empire's atrocities, and the fate of the characters introduced in the previous novels. What's interesting is the recurrent character General Jedao, who's been cloned, transplanted, and brought back to life. We also get to see that the star-ships used by the civilizations in the novel aren't really manufactured, but are living organisms that are grown and then harnessed.

The biggest failure of the story is that the ideas behind main plot point (exotic technologies that can only be used in terrain that subscribes to a certain calendar) aren't fully fleshed out: you could easily file all the numbers off and replace technology with magic and the story would still work.

On the other hand, the story reads easily, the characters sympathetic (except for the villain, who's also never fleshed out enough for you to have any sympathy for him), and the payoff, such as it is, has plenty of action and explosions. It's a light airplane read that serves its purpose. Mildly recommended.

Monday, September 23, 2019

Review Keychron K2 Bluetooth Keyboard

I ordered the Keychron K2 Bluetooth Keyboard as a kickstarter project for a couple of reasons:

  1. I wanted a wireless keyboard, and the K2 advertised itself as being pairable with up to 3 computers, so you could pair it with desktop, personal laptop, and say, work computer.
  2. My wife kept complaining about how loud the Varmilo VA87M was. The keyboard had a great feel, so I knew I wanted another mechanical keyboard.
With kickstarter, you never know when or if the keyboard would ever show up, but I figured since Keychron had actual products on Amazon, including the K1, I figured it wasn't too much of a risk. I didn't look too carefully at the layout of the original K1 vs the K2, or I might have decided that the extra thinness of the K1 was worth it in exchange for better layout.

When the keyboard arrived, I was impressed by how compact and quiet it was given that it had all the arrow keys and the "home" and "end" buttons. What was missing is the "insert" button, which I do miss once in a while. The keyboard came with a keypuller and extra keycaps in case it bothered you to have the device keys look like a Mac keyboard instead of a real Windows keyboard. Since I mostly use windows on my home systems I switched out the keycaps.

Pairing is quick and easy: hold down Fn and one of either '1', '2', or '3', depending on which slot you want to pair with. The device paired easily, and had a neat little feature, which is that as you type, the keys you're typing on light up and then fade slowly away. Very cool! It's a frivolous feature but it really tickled me.

The problem with bluetooth keyboards of course is that they're not as responsive or as low latency as a real keyboard hooked up to the machine. Fortunately, the device comes with a USB slot that you can plug in to either charge the keyboard or to get a real time response when plugged in instead. When plugged in, I expect that the latency and response time will be just as good as a real wired keyboard.

The bluetooth stacks on computers, phones and keyboards all suck: after a long time in disconnected state, the keyboard will no longer re-connect with my desktop. I can't tell whether it's the cheap CSR dongle that I got off Amazon for $6.70, or this keyboard. I suspect a more modern bluetooth 5.0 dongle with better firmware (I since tested the keyboard against a MacBook Pro, and indeed both the pairing/resumption from sleep functions work much better, so I guess I should bite the bullet and buy a higher quality BT dongle!) will work better, but since re-pairing the keyboard is easy, I just live with it. (I have a Logitech Brio so I can login even without the keyboard)

The keyboard has a very nice typing feel. It's a little software than the Varmilo, but still much better than the crappy chiclet/island-style keyboards that come with more modern computers, and I can maintain a high typing speed on it without errors, provided the bluetooth latency isn't a problem.

To my surprise, the keyboard battery ran low after just about 4 weeks of use. Then I realized that I'd inadvertently left the keyboard's lighting mode on. I suspect that keeping the keyboard's backlight off would grant you better battery life. In any case, the keyboard is only rated for 15 hours of use, which indicates that bluetooth is a much more power expensive protocol than whatever it is that Logitech's unifying technology uses.

All in all, for the price it's a great keyboard. If you need a bluetooth keyboard that can also be a wired keyboard, this is the only one that'll take USB-C, so you don't currently have a lot of other choices! That it's fairly cheap makes it unique in its class. Recommended.

Friday, September 20, 2019

Review: Raven Stratagem

Raven Stratagem is the second book in Yoon Ha Lee's space opera trilogy.  The previous novel didn't end exactly on a cliffhanger, but clearly the plot had started going and there were many loose ends left untied. This one begins with a deception, though one that careful readers of the previous books could easily see.

As with the previous book, there's precious little science in this space opera: it's not really science fiction, and it's more of a character study than it is say, a novel of political intrigue (though there's quite a bit of it). I enjoyed it as a diversion, though it's nowhere as good as say, early Richard K Morgan. Recommended.

Thursday, September 19, 2019

Review: The Sociopath Next Door

I picked up The Sociopath Next Door during an audible sale. The biggest issue with audio books is that the voice actors reading the book might not be appropriate for the book's topic. This is particularly the case with this book: the actress reading the book read in a calm, gentle, languorous voice, completely at odds with the topic of the book.

The book, obviously, is about sociopaths, but works hard to debunk the typical impression you might have from the media, which is that they're the Hannibal Lector crime committer or serial killer. The reality is, the definition of a sociopath (according to this book), is that the person has no conscience, need, or ability to empathize with other human beings. That means they're free to do anything and everything, and will treat other people as objects or obstacles to achieving their goals.

As you can guess by now, most high level executives and CEOs are sociopaths. And of course, the profit-driven charter of most corporations means that by law, they are require to behave as though they're sociopaths! The book runs through several case studies of actual sociopaths and how they may behave with cruelty towards their peers, their patients, their students, and even their families. Then it goes on to describe the author's theory about how sociopaths may have evolved genetically, and why there are so many more sociopaths in Western societies than there are in Asian societies. This part of the book feels a little iffy --- maybe in the paper or Kindle version of the book there are references to research, but the statistics are that something like 4% of American society are sociopaths, and about 0.5% of Asian societies are sociopaths. There's some speculation about why this is so.

There's a tiny, weak chapter about how to deal with sociopaths if you find yourself in the path of one. I didn't think it had outstanding advice. In any case, the book's short, covers an important topics, especially if the 4% figure is true, and is worth your time. Just get the paperback version unless you enjoy sleeping to the sound of this particular actress's voice.

Wednesday, September 18, 2019

Review: Nine Fox Gambit

Nine Fox Gambit is Yoon Ha Lee's first book in a space opera trilogy. While it has all the trappings of science fiction, it's actually a fantasy story, with magic powers and a magic system based on a complex calendar with much of the elements of the traditional Chinese lunar calendar (Lee is Korean American). The calendar system determines the powers of various battle formations, when it would afford attacks, and even what kinds of weapons are viable in a space empire.

The magic system is never fully explained: you're expected to infer it from the conversations and the actions of the protagonists and antagonists in the novel. A major plot point comes from a rebellion within an empire where the rebels have chosen a new calendar system as well as a new form of government (a democracy), and the reader is treated to a view of democracy as being a "known to have failed" system.

It's a rollicking read, and very compelling: I found myself looking forward to picking up the kindle every evening to learn more about the characters and situations. Recommended. I'm proceeding right to the next book in the series.
First: the value of a game is in abstraction. Many Nirai go in for simulationist approaches, a tendency you share, but sometimes you learn more by throwing details out than coding them all in. You want to get rid of everything nonessential, cook it down to its simplest possible form.” (Kindle Loc 3629)
 “According to the Shuos,” Jedao said, “games are about behavior modification. The rules constrain some behaviors and reward others. Of course, people cheat, and there are consequences around that, too, so implicit rules and social context are just as important. Meaningless cards, tokens, and symbols become invested with value and significance in the world of the game. In a sense, all calendrical war is a game between competing sets of rules, fueled by the coherence of our beliefs. To win a calendrical war, you have to understand how game systems work.” (Kinlde Loc 2640)

Tuesday, September 17, 2019

Review: The Sympathizer

The Sympathizer is Viet Thanh Nguyen's novel of the Vietnamese-American experience. In particular, it could only have been written by a Vietnamese-American who's grown up with the English language with eagerness: his use of language is erudite, eloquent, and nuanced, while his exposition of the Vietnamese-American experience itself can only have been written by someone whose command of the English language betters that of most native English speakers, but is still nevertheless rejected by the mainstream culture by virtue of his cultural background, skin color, and of course, his history.

The book won the Pulitzer Prize, and generally books that win such prizes are not much fun to read, being concerned with being in the vanguard of Literature and Literary achievement rather than story, character, plot, and being interesting. For most of the novel, Nguyen succeeds in avoiding such idiocy. The story revolves around a Eurasian man of descent who decides to be a double-agent, working with the communists to undermine the South Vietnamese interests. That experience, of course, is a great analogue with the Vietnamese American experience: someone who sympathizes with both the Vietnamese cultural experience and the American culture he finds himself immersed in.

Thus it is that we see great scenes, including the one where he, as a cultural expert, is involved in one of the many American movies written for (and by) Americans about the Vietnam war, but which of course, dehumanizes the Vietnamese as much as possible while still putting Americans (even American villains) on the center stage. He tries to change the movie by working from the inside, but of course fails.

Lots of similar scenes recur, and the story does cover the experience of being a refugee. Where the book flails is when Nguyen writes himself into a corner, by having the protagonist return to Vietnam. There's no reasonable setup in which this can happen and the protagonist has an exit (especially one in which he gets to write down what happened), so this section degenerates into magical realism and fantasy land.

Nevertheless, I found the novel worth reading. Recommended.

Monday, September 16, 2019

Review: Your Best Brain

Your Best Brain is by John Medina, since I'm a big fan of his, I picked it up despite knowing that it covers little new ground compared to the books I'd already read, e.g. Brain Rules.

I'd say that the biggest difference between this "Great Courses" audio lecture series and the books is the emphasis on social interaction. Medina emphasizes how important that is to longevity and brain function, which I don't remember him emphasizing in his previous books. Each lecture is accompanied by a section on "practical advice", but he admits that they're still few and far between.

If you haven't read any of his previous books, this lecture series is worth listening to. If you have read his previous books, this will serve only as a review. Recommended.

Friday, September 13, 2019

Review: Dopesick

Dopesick is a book exploring the opoid epidemic in the USA, specifically in Virginia. Scarlet pointed me at the book, and it's very much worth reading for an overview of what some economists call "The diseases of despair." I learned a ton in this book.

For instance, I didn't know that selling addictive drugs has long been part of the history of the big pharmaceutical companies:
By 1899, Bayer was cranking out a ton of heroin a year and selling it in twenty-three countries. In the United States, cough drops and even baby-soothing syrups were laced with heroin, ballyhooed at a time when typical opioid consumers were by now not only war veterans but also middle-aged barbers and teachers, shopkeepers and housewives. Many were mostly functioning, doctor-approved users, able to hide their habits—as long as their supply remained steady, and as long as they didn’t overdo. (pg 24)
Even more importantly, I didn't realize that the subjective pain management scale and the relatively recent emphasis on pain management was propaganda foisted onto the medical profession by the pharmaceutical companies. It's astounding to me that the FDA allowed opoid-based pain medication out along without guidance on the possibilities of addiction to powerful drugs. Many of the drug addicts became addicted not because they were partying, but because they were prescribed a drug for too long (and it's not clear to me how studies are done about potential addiction to a pain-killer --- I remember a nurse injecting morphine into me after I'd gotten hit by a car, telling me: "don't worry, you won't get addicted to it." Mercifully, right after that injection I lost consciousness, and needless to say she was right --- I had no craving for morphine after that).
New patients were given OxyContin “starter coupons” for free prescriptions—redeemable for a thirty-day supply—and Purdue conducted more than forty national pain management and speaker-training conferences, luring doctors to resorts from Boca Raton, Florida, to Scottsdale, Arizona. The trips were free, including beach hats with the royal-blue OxyContin logo. More than five thousand doctors, nurses, and pharmacists attended the conferences during the drug’s first five years—all expenses paid. (pg 47)
Again, this is only something that happens in the USA. Other countries with government provided healthcare wouldn't have marketing driven prescriptions. Again and again, the politics of the situation rears its head. For instance, Virginia was one of the states that fought Obamacare by not expanding medicaid coverage.  That creates a cycle of drug use, since many of the addicts were too poor to afford rehab (not that they would have wanted rehab --- many of the victims had families that paid for and pushed them into rehab). After a US district attorney successfully sued Purdue Pharmaceuticals and got them to cough up a relatively small settlement, he was almost fired:
Eight days after it accepted the deal, Brownlee was stunned to see his name on a firing list, along with four other U.S. attorneys. Though he wasn’t ultimately fired, the incident provided fresh criticism of then–attorney general Alberto Gonzales, accused of trying to sway the work of U.S. attorneys’ offices. And it only underscored the long reach of Purdue: Udell’s defense lawyer Mary Jo White, a former Manhattan U.S. attorney, had been the one to press for more time in a call to a Department of Justice official. (Brownlee would break down how Purdue’s attempted influence peddling worked—or didn’t—in a later Senate hearing about the case.) (pg. 82)
I'm always amazed by how Americans seem to fear the government, which is subject to elections and public oversight, while heaping social approval on corporations, who are effectively legally required to be psychopaths in pursuit of maximum profits regardless of the social consequences, can't go to jail, and have so much money that even multi-million dollar fines have no incentive effect on their  behavior.

More importantly, the book provides case studies of the drug addicts. This not just puts a face on the victims, but also points out that many of these drug addicts do not fit your stereotype of one: these were successful men and women from good families who were doing well in school. Drug addiction really does change your brain, and turns you into a desperate junkie in search of the next fix no matter your economic class or social status. (There's even a person who was a doctor on probation due to addiction) If you're a parent you will probably come away from this book fearful for your kids.

Ironically, these diseases of despair only hit white people. The black population was spared this because:
Why had blacks failed to become ensnared in opioid addiction? That question was addressed in 2014 data issued by the Centers for Disease Control and Prevention: Doctors didn’t trust people of color not to abuse opioids, so they prescribed them painkillers at far lower rates than they did whites. “It’s a case where racial stereotyping actually seems to be having a protective effect,” marveled researcher Dr. Andrew Kolodny of Brandeis University. Put another way: By 2014, while young whites were dying of overdose at a rate three times higher than they did in 2002, the death rate for people of color was relatively unchanged. (pg. 253)
There's lots more in this book, such as the political fight over effective rehab (the scientifically proven to work stuff isn't used because the religious people and the AA types are ideologically opposed to it), and of course the individual case studies.
—he believes the five-year treatment model, common for addicted doctors and airline pilots, is ideal. It’s why they tend to have opioid-recovery rates as high as 70 to 90 percent. “There’s nothing scientific at all about twenty-eight days of [residential] treatment,” Loyd said of the kind heralded in movies and on reality TV. “It takes the frontal lobe, the insight and judgment part that’s been shut down by continued drug use, at least ninety days just to start to come back online and sometimes two years to be fully functioning.” (pg. 294)
There's a poignant section of the book where the author attends a "free healthcare" camp and compares it to 3rd world countries like Haiti:
In rural America, where overdose rates are still 50 percent higher than in urban areas, the Third World disaster imagery is apt, although the state of health of RAM patients was actually far worse. “In Central America, they’re eating beans and rice and walking everywhere,” a volunteer doctor told the New York Times reporter sent to cover the event. “They’re not drinking Mountain Dew and eating candy. They’re not having an epidemic of obesity and diabetes and lung cancer.” (pg. 274)
There are comparisons with other countries that have decriminalized drugs and reduced the cost of addicts to society by providing working rehab programs. Again, ideology (mostly from the non-evidence based crowd) trumps effective operation of society in the USA in a way that's uniquely dysfunctional.

While the author works hard to draw your sympathy, by the end of the book I was starting to get irritated that the voters of Virginia continue to vote for politicians that propound ineffective solutions that kill their children and loved ones while criminalizing them and blaming them for being victims of big pharma. Unfortunately, this book does not have a happy ending, and of course, it's not clear that the country as a whole will have a happy ending if voters keep operating as though the opoid epidemic is a crime issue rather than a medical issue.

The book does have weaknesses: I would have wanted to see a more involved examination of the various rehab approaches and the studies involved. I would rather have seen more statistics about drug movement and distribution rather than a lone interview with a drug dealer (who, like the executives at Purdue Pharmaceuticals, demonstrated no remorse about the lives he ruined). But overall, that does not detract from the point that the book's very much worth your time reading. Recommended.

 
 

Thursday, September 12, 2019

Review: Prognosis

Prognosis was for me that rarest of books: an Amazon First giveaway that drew me in and force me to read until I got to the finish. The author, Sarah Vallance was a smart, risk-seeking woman who was thrown off a horse and hit her head. This resulted in what's called a "mild brain injury."

As one of my friends learned ages ago, there's no such thing as a mild brain injury, and Vallance describes her process of recovery and healing from the trauma well: from the horror of not having executive function to months of recovering the ability to read or even put together sentences, and the stories of the medical professionals writing her off after she scored with an IQ of 80.

Part of it was her very angry personality, deliberately refusing the help of social workers and other professionals. On the other hand, this was someone who'd suffered a brain injury, and the professionals should know that brain injury victims are literally unable to make rational decisions. On the other hand, she was very lucky to get her brain injury in a country with universal healthcare. In the USA, she'd probably have had her house taken away, and ended up homeless and there would be no way that this book would have been written.

The recovery process was painful to even read, but what came through in the end was that despite her self-doubts, Vallance made a good recovery and performed at a very high level after she attained her PhD, won a fellowship, and became a HR professional and expat. While she's not a very likable person (and neither were much of her family as described), the writing flows, is easy to read, and you will have sympathy for her by the end of the story.

Recommended.

Wednesday, September 11, 2019

Review: Verbal Judo - The Gentle Art of Persuasion

Verbal Judo's subtitle is The Gentle Art of Persuasion. To my surprise, the book's written by an ex-cop.  In one sense, this is good. You want tried and true techniques that have been used in life-or-death situations and you hope that's generally applicable to your daily life. On the other hand, it's not quite clear that some of the techniques he described would be ones you would (or could) use in domestic environments.

For instance, in one instance he's trying to talk down a crazy guy from wounding his son because he thought his son was possessed by demons. His solution was to lie to the guy and said: "I know a priest specializing in demonic possession." Of course in this case it worked, and he was able to save a child, but I'm pretty sure if you lied to your wife or kids about anything you'd be caught sooner or later, and that's going to ruin your credibility. Since cop-citizen interactions are frequently one-off interactions, the issues that come into play in domestic life don't come into play that often. Similarly, a cop making an arrest has formal authority and power that you probably don't have most of the time when dealing with other people in business.

That said, the book has a ton of advice that is generally applicable, including how to take a step back and keep yourself from saying things you'll regret later. That self-discipline is a useful one, and I think more people could benefit from it. Similarly, the approach towards summarizing the situation is a good one.

As with many books of this nature, it's frequently repetitive and over-written. On paper or on the Kindle, this is a big deal, but since I audited the audio book version, the excessive repetition was fine since in an audio book you can't go back and re-read stuff that's referred to earlier, so in some sense that's an advantage. Recommended.

Tuesday, September 10, 2019

Review: Instructions

Neil Gaiman and Charles Vess have been collaborating on a series of children's books. I didn't know about them until I saw Instructions on the used books section of the library. For $2, I thought that it was a cheap risk.

While Bowen was happy to have me read to him when he was younger, Boen's always been a different story, never wanting to read books. Even when he was willing to read books, he'd prefer stuff like Monster Trucks. But as an experiment I read Instructions to him and he immediately asked to have me read it to him again!

Unlike many other children's books, there's no immediate morals, or even much of a story line that's fully fleshed out in this book. Mostly, it's hints of a story (fed by instructions), leaving the reader to tease out the characters and events through illustrations. Even for adults, this is intriguing and interesting. The language and artwork complement each other, and of course, it's a very short read (10 minutes max).

Recommended! I'll go look for other children's books by Gaiman and Vess.

Monday, September 09, 2019

Review: Samsung Galaxy S9+ (128GB storage, 6GB of RAM)

This year's trip finally convinced me that in 2019, it is insufficient to have an Android phone with only 3GB of RAM. On the days when I was using Google Maps to navigate the route, the any attempts to take pictures with the phone would cause my Moto X4 to stop navigating. And of course, when my Canon G7II failed, my phone became my backup phone. I don't expect much from phone cameras: the idea is to take pictures that are better than nothing. The problem with the Moto X4 was that the camera startup time was 5-7s. That's an eon. By contrast, my wife's Pixel 3a XL always kicked off the camera quickly, and of course, the camera on that phone takes superlative pictures.

During this year's Prime day, the Pixel 3a XL was sold with a $100 coupon attached to it. This was tempting, but I came to the conclusion that the 3a XL was too much of a compromise: it wasn't waterproof, didn't support SD cards, and came with only 64GB of storage. I could probably give up any one of these in exchange for the pure android experience and superlative camera, but all 3? That was too much. There were also a couple of other niggling things, such as the fact that the Pixel 3a XL is not compatible with QC chargers (it only charges quickly with the USB PD standard). Google's product managers will never learn that if you're not Apple, you can't unilaterally impose a new standard on customers without that annoyance turning off their customers. Whoever's sitting in Android HQ thinking that customers just want an iPhone, only running Android OS is smoking some really good stuff.

By contrast, Arturo's Samsung Galaxy S8 had been used for the last couple of years with no problems. He'd even cracked the screen and jumped into a lake with the same phone and it was still waterproof. Arturo is almost as tough on equipment as I am, and uses his phone way more than I do, so I reluctantly came to the conclusion that if he could come to terms with the Samsung bloatware, I could do the same. He also claimed that a software update in the middle of the lifecycle actually improved the battery life of his phone! It helped that on Prime day, the Samsung Galaxy S9+ with 128GB of storage and 4GB of RAM was $500. I considered other phones, such as the Galaxy S10 (similar specs, but with a 3rd camera, in-screen camera reader, $100 more, worse battery life), and the Moto Z4 (not waterproof, worse everything else, same price), and decided that the S9+ was the best fit: Samsung had fixed the problematic fingerprint placement on the S8, and the additional performance of the S10 didn't justify $100 more. I was pretty sure that I didn't want a wider angle lens than the one that came with the S9+. Note that the S9+ 64GB would have only come with 4GB of RAM, and I decided that if 3GB was sufficient in 2016 but not in 2019, I'd better have more than 4GB.

To begin with, the phone is fast. The camera starts up with a double-click of the power button in a second, as fast as my wife's Pixel 3a XL. You can then use the volume down button to shoot pictures, or you can reprogram the Bixby button to do the same with a $3 paid app. In practice, the phone is still more awkward than the 3a XL because the volume/bixby button is opposite from the power button, rather than next to it, but I could still shoot reasonable pictures while cycling:
Having said that, I'm still not as competent as on the G7X II with its real shutter buttons --- I've shot many pictures of my stokers on the back of the tandem shooting blind, but I tried the same on my S9+ and couldn't get a single usable shot! However, it did take me a while to get used to the G7X II's blind shooting capabilities as well, so there's a good chance that in a couple of months I'll be as good with the S9+ as I was with the G7X2.

There's a 2X zoom on the camera, and I thought I wouldn't use it, but in practice once you know it's there, you use it a surprising amount, which makes me think that the ultra-wide camera on the S10/S10+ might not be a waste at all.

Arturo complains that the plus size version Samsung's phones are too big, but when I compared it with my wife's Google 3a XL phone, they're about the same size, and because of the smaller bezels, the S9+ has a bigger screen. Compared to my older Moto X4, it's about 1cm longer, which isn't enough to cause it to fall out of my favorite Columbia Trail Splash shorts, even with vigorous cycling. Screen size is a big deal: for viewing maps, photos, movies, and web pages, you want as big a screen as you can get. The only thing I'd complain about is that I think they could have made the screen narrower to make it easier to reach the entire width of the screen with one hand. But I'm not going to complain about the ergonomics --- it's still better to have a bigger screen.

The bloatware is annoying, with 2 of everything, including calendar, e-mail, and even web browsers. I just disabled everything I found annoying, but of course no doubt some of the performance of the phone is compromised because it's running 2 of everything. The audio is great, and of course, unlike the Moto X4, the S9+ supports band 71. Until 5G becomes ubiquitous, this is a good setup. It's actually the only justifiable reason to get a flagship phone: excellent world wide coverage with better antennas.

There's an amazing number of features on the phone, including some that I'll never used, such as support for Ant+. (I have bike computers for hat) Similarly, there's an O2 sat sensor that I used once and decided it wasn't worth the effort. I looked for an infra-red port but there wasn't one, and of course, there's an FM radio that you can use where your wired headphones become an antenna, that's got great reception, something none of the flagship phones from Apple or Google have. This could turn out to be a life-saver in emergencies when cell phone towers go out. Since nobody has a wired landline any more, I think it's worth shopping around for a phone with this feature. (And yes, the idiot lemming manufacturers who followed Cupertino into killing the headphone jacks don't offer this, so Samsung, LG, and Motorola are pretty much the only phones left with this feature --- even the 3a XL doesn't offer this despite having a headphone jack)

The extra memory was disappointing: I hoped that the extra memory would be put to use and my commonly used apps like Audible audiobooks wouldn't get swapped out. Nope. Audible still occasionally gets swapped out even as the amount of "free memory" in the status pages is north of 1GB. I couldn't help wishing for more options to say, pin certain apps in RAM (hey, I've got plenty of it, why not give me some control) so they don't ever get closed or swapped out.

What surprised me was how much better the bluetooth audio was: my Taotronics headset, despite being too soft with the Moto X4, could be tuned louder with the S9+. I really don't understand why. The headset doesn't draw power from the handset, so there's no reason why one phone should produce louder sounds than the other, but for whatever reason, the Samsung definitely sounds clearer.

The NFC chip is in the middle of the phone, rather than the top edge where the Moto X4 was. I had to retrain my muscle memory to use it. Samsung Pay, of course, is superior to Google Pay, mostly because of the hardware. Even if there's no NFC chip on the transaction terminal, you can use Samsung Pay because the device works with magnetic card readers as well as NFC terminals. In previous years when NFC terminals weren't prevalent this would have been a major feature. Now it's just "nice to have", since NFC payments have been made available nearly everywhere except the post office and certain government offices.

Phone calls are crystal clear and a noticeable improvement over the Moto X4. Most users probably don't care, in this day and age.

Battery life is good in default mode. There are various battery management modes that I didn't know until Arturo told me during a recent backpacking trip. In the most optimized mode, I went from a 99% charge to 74% during a 36 hour overnight backpacking trip with plenty of picture taking, bluetooth turned on for connection to both my Fenix 5X and the AirMini, and occasional use as a flash light, etc. (Airplane mode was on)  This is remarkable battery life and a huge win. I'm very impressed, because I didn't notice any missing functionality and didn't baby the phone at all.

I actually tried Samsung Dex for half a day. I was very impressed. You can plug your device into a standard usb-C dock, and get 1080p resolution on a big monitor, a wired keyboard, and bluetooth mouse pairing. To get 4K resolution you need to shell out for the $90 official Samsung adapter (https://www.samsung.com/us/mobile/mobile-accessories/phones/multi-port-usb-c-adapter--black-ee-p5000bbegww/). While not all applications happily adapt to a desktop environment, ConnectBot (https://play.google.com/store/apps/details?id=org.connectbot) does, and grants you a reasonably good ssh experience, complete with cut/and paste. If you're a C/C++/Go programmer, Termux (https://play.google.com/store/apps/details?id=com.termux) even lets you develop in a semi-linux environment right on the phone, but doesn't support proper cut/paste.

Unlike Google phones, Samsung's phones do not get updated frequently. Even worse, the updates are frequently gated by your carrier, even for unlocked phones. I'm past the point where "upgrades" have a positive connotation --- the constant UI changes (probably driven by UI designers in search of a promotion rather than any desire to actually improve the user experience), so this doesn't bother me, but it might bother you.

Would I have paid the full $840 price for the S9+ when it first came out last year? No way. But at $340 less for the upgraded 128GB/6GB RAM version? I think it's well worth the slight premium over the Pixel 3a XL. Waterproofing, microsd expansion, QC compatibility, and more RAM are well worth the extra money, no matter what Google's product managers think.

Friday, September 06, 2019

Review: Purdoux Cotton CPAP Mask Wipes

When it comes to stuff like CPAP mask wipes, I'm discovering that the most important factor is the packaging. You can get those big cylinders of wipes,  but those are painful to bring on a bicycle tour, camping, or even a sailing trip where you have to start by getting onto a plane. Then Lofta sold me some Purdoux Mask wipes.

These are the perfect packaging. They come in satchels of 10 wipes each, so you can bring just the right amount for your trip. (For a 3 week trip, bring 2 satchels and then just reuse the wipe for the last day) They're resealable, but once you've opened a satchel to prevent them from drying out just use up the entire satchel. They're more expensive than the big cylinder wipes, but if you have a habit of going on bike tours or trips where light weight is essential, these are a great supplement. Recommended.

Thursday, September 05, 2019

Connection Options for the LG 43UD79-B

Over the last 6 months, I still have not found a monitor as good as the LG 43UD79-B.  It's huge and does not require font scaling at 4K, and performs well for writing, photography, and video editing. I feel like I'm going blind every time I have to use anything smaller.

The monitor, however, comes with 4 HDMI ports and only 1 display port as well as a single USB C port. 2 of the HDMI ports are special, allowing for 4K60fps. While the display port is reliable and has always worked, as soon as I acquired the XPS 13, I wanted to be able to connect it with the monitor. This demonstrated that several of the cabling standards are anything but standard or reliable, so I'm documenting everything I tried.

The first thing I tried was a Keychron 10 in 1 USB Hub adapter.  This one was very appealing, because it was cheap (about $40 if you find a coupon), had 4K output, charge the laptop, as well as providing 3 USB hubs and an ethernet jack as well as a VGA port. (nowadays, monitors don't even come with a VGA port, but you might have a legacy monitor sitting around) The problem? No 4K60fps. 4K30fps max. Back it went to Amazon.

Then I tried the Cable Matters USB C with DHMI, DisplayPort, VGA, and Ethernet adapter.  No HDMI 4K60fps either, even though the very same adapter's Display Port would give you 4K60fps. But the DP port was already spoken for by the desktop machine, and seriously, if I was going to do that, why not get the Amazon Basics adapter, which is way cheaper?

Finally, the uni USB C to HDMI 4K60fps cable worked. I went for the short 3 foot cable, and this one was reliable, and gave me 60fps. At $19, it's not the cheapest cable in the world, but as I've already discovered not all HDMI cables are made equal, so if you need one of these, this is the one to get.

I'd tried several USB-C cables that claim to be alt-mode compatible. This has great appeal, since the cable has the potential to provide USB, display output, as well as charge the laptop, simplifying your setup to only one cable. To my surprise after the previous years of trying one cable after another and failing, the cheap ($12) Nekteck USB C 3.1 Gen 2 cable worked, delivering me 60fps.  To my disappointment, the XPS 13 wouldn't charge from the monitor, as the monitor only delivered 7 watts of power!

So there you go, I've got 2 cables that'll work with 4K60fps, and 2 that wouldn't! Compared to last year, the situation is definitely improving!

Wednesday, September 04, 2019

Review: The Weather Machine

The Weather Machine is a history of the weather forecasting system. It discusses a history of weather forecasting, the network of weather stations that collect the data, the birth of the satellite weather observation stations, and the rise of modern super-computers that do the forecasts that have become fairly reliable in the short term (1-3 days).

It's a short read, and I love short reads that provide deep insight, but unfortunately in this case the coverage is extremely shallow. There's no coverage of how the forecasting computation actually works. There's no coverage of how the new systems of climate attribution work. There's nothing about how the old "front" system of forecasting works.

In other words, this is a "science book" written by English majors for English majors who don't actually know anything about anything. The author probably got excited about this book in the morning, whipped up a short summary by afternoon, and got bored 2 weeks into writing the book and couldn't be bothered to actually turn it into a book with real substance. I ended the book disappointed and was glad I didn't pay money for it.

Tuesday, September 03, 2019

Review: The Story of Human Language

The Story of Human Language is one of the best "great courses" audio lecture series I've ever audited. I have no formal training in linguistics other than auditing other lecture series, and came into the lecture series thinking it'd be something like a history of the English language, but instead it's a principles-oriented discussion of language evolution and the implications of those principles!

Here are the principles as discussed:

  • Vowels at the end of words are in danger of being eroded away
  • Vowels tend to shift over time
  • Words and phrases tend to get shortened and sped up over time
  • Intensity of semantics tends to erode overtime, so meanings get diluted. (Think about how you used to be able to say "Great!" and mean it, and now you have to say "Awesome")


    From these, John McWhorter discusses:
    • Where do tones come from? (if you erode away words, eventually what's left as the difference between words becomes tonal)
    • Are all languages equally complex? (No, the complicated languages are the ones that are spoken by a small insular group of people with no larger interaction with the outside world --- when adults have to learn a language they learn it badly so complicated languages get shaved down and simplified... As a result, the more widely spoken a language is, the simpler it becomes --- as a result, Mandarin Chinese is easier to learn than Cantonese, for instance)
    • The corollary is that a language that grows out of a pidgin gradually develops grammar and becomes more complicated as it needs to express more, but is still easier that ancient languages.
    • Because words tend to disappear over time or get diluted, only written languages can maintain a big vocabulary.
    There's a ton of other stuff in here, including a discussion of what's a language versus what's merely a pidgin (this is highly technical). It turns out that Hawaiian Pidgin is actually a creole language. There's a great explanation of why English is so different from old English, and why we find Shakespeare nearly incomprehensible. There's a discussion of artificial languages such as Esperanto and the various sign languages including Solresol, a musical constructed language. It sounds so delightful that I had to look it up after auditing the entire course.

    I can't do this audio lecture series justice. Just the anecdote about how he walked into an elevator with two men who grew up together but hadn't seen each other in decades, and how they gradually code-shifted from formal English into their native dialects made the entire series worthwhile. It's well worth the 18 hours (or you could speed it up with the audible app by listening at double speed) . This is the "great courses" audio lecture series where I went back and read the lecture notes that accompanied the audible download.

    Highly recommended!