For his own health I am glad that he didn't continue the diet much longer. There are a number of confounding variables here that the article forgot to mention, which most likely played a larger part in his weight loss than any calorie counting:
Stress. Stress can cause sudden weight loss. This weight will come right back if you give your body any time to catch up and yet the effect is significant enough to provide "evidence" for the craziest of dieting fads. Eating debbie cakes for 3 meals a day is going to put some serious stress on your body.
Metabolism is not constant. Michael Phelps eats more for breakfast than I do in a week, and yet he is in much better shape. The scary thing about junk food is not so much the empty calories, but more how it affects your metabolism in the long term. If you draw out this sort of a diet for 2 or 3 years, then I would expect that not only could you give yourself diabetes, but also you will start craving more and needing less. That is when weight gain will start to become a serious problem.
There are a number of confounding variables here that the article forgot to mention, which most likely played a larger part in his weight loss than any calorie counting.
I disagree completely. Stress and metabolism are minor factors compared to the repeatable, testable, and well understood calculus of calories in vs calories out.
I'm sorry but in the Health and fitness community this has been disproven time and time again. Someone can fix their sleep and eat at their bmr with good foods and still lose heaps of weight. See good calories bad calories by Gary taubes
Basal metabolic rate is something developed in hospitals for bed-ridden patients on total parenteral nutrition. It is the calories you burn just staying alive. If you, a healthy person walking around, eats at your BMR, you will, definitely, loose a heap of weight. However, the BMR for the average human is quite low.
Also, on brief review, Taubes appear to thoroughly misunderstand the issue on a variety of levels. I'm going through more of his material, if I am mistaken, I will report my error.
I don't disagree completely. We are talking about completely different things.
I am talking about long term, you are talking about 10 weeks. Yes, you can calculate calories in vs calories out. No, you cannot extrapolate that to determine if Professor Mark Haub will be alive on that diet in 3 years. What is the metabolism of dead?
What repeatable calculus? This is a story about one guy not a study of a population of people. I'd be cautious before you jump to the conclusion that these results will generalize. That probably says more about your own biases than it does any scientific learning that was produced by this experiment.
The Harris-Benedict equations are generally accepted as empirically corroborated and reliable predictors of weight change as a function of daily caloric intake.
Of course. Most dieters don't have access to a ventilated hood indirect colorimeter, so they have to rely on a method for estimating their BMR.
What I'm saying is that the difference between their measured BMR and estimated BMR (using either Harris-Benedict, WHO, or Schofield equations) is a minor factor.
I'm saying that I believe that diet has a real effect on metabolism. Not basal metabolism, but on the ability to maintain a given activity level. If you try and restrict your calorie intake, you will be less likely to go on a run, take a walk - etc. This is why Harris-Benedict may not be a good tool to manage long term weight loss, even though it is a useful predictive tool.
I am not sure how applicable it is to an introductory text, but I would be interested in learning more about the internals of compilation and linking. What is the minimal necessary amount of work that needs to be done to interact with existing C libraries (header files, ELF format, and how that all fits together).
This keeps a playlist of what you liked and uses that to recommend videos, allowing you to personalize and filter down the videos to only what you are interested in.
Example: Searching for "haskell" will give you videos about people named haskell as well as the programming language. After adding a few example videos to your list, you will get only results related to the programming language. The idea is to start to combat the incredible amount of noise in search results: believe it or not, not everyone enjoys rick rolls and gophers.
His advice was that "you always focus on supply first... then you shift to demand." Here that would mean to try to grab producers before promoters.
The idea is that producers are more willing to list their movies on your site, no harm done, but promoters won't be interested unless you have some movies for them to choose from.
From Startup School as well, Adam D'Angelo said they answered a lot of questions themselves at first, Mark Zuckerberg said they had interns scrape the course info manually, Brian Chesky talked about how they recruited the first people themselves by door-to-door basically.
Summed up by D'Angelo: "it doesn't matter if it doesn't scale if it strengthens your position". Though I don't know how that applies to your case, if there's anything you can do to get things started.
2cts I think he said go to the harder side 1st (if it's supply, supply, if it's demand, demand) and then shift to the easier side. For me was the best talk from Startup School 2010 with its story: "Go to your users"!
After having you point out that fact, I had to do a search on the document source for "<a" to find the 6 point font footer at the bottom! I hardly expect anyone to do this much work looking through my job applications...
For me, self submissions are a large part of the reason why Hacker News is interesting. A lot of smart people come here to share their thoughts. Without that we would only have the Times blogs which aren't hard to find in the first place.
I am by no means representative of the Hacker News community, but I would guess that other people would say something similar.
I've always thought of dynamic programming as memoized computations + ordering. Maybe memoization isn't quite as efficient as dynamic programming but it feels like it is halfway there.
Good point. Memoization is a one off that I also find myself using quite regularly.
"Coolest" algo I've ever developed would have to be parser combinators. Coolest graph algorithm: finding maximal cliques and independent sets + approximation algorithms for these.
It is only possible to measure fitness for a specific task. Often a trait that would help in one situation will hinder in another.