Monday, January 29, 2007

Video Server

I have a wireless Linux fileserver that automatically scans all binary newsgroups for rotating keywords of my choice (while I’m at work), then downloads found DVD images onto its homemade RAID-array, after which it wirelessly shoots combined video to my new 42″ tube and audio to the Soundblaster X-Fi-driven 7.1 surround sound speaker rig. I control the server with a wireless mouse and a small LCD display. I do not own a DVD player.

Sunday, January 28, 2007

Verizon Spends Half-Billion to Connect U.S. and China

Verizon this week announced a major project that would connect the U.S. and China together for the first time through a high-speed fiber optic line. The project is designed to provide more bandwidth between the two major countries, especially when network communications is so critical to businesses and consumers these days.

The new 11,000-mile fiber line will cross the Pacific Ocean and co-exist with a current system that has reached its limits. According to Verizon, the new line will have more than 60-times the bandwidth capacity of the current system allowing as much as 62-million simultaneous high-quality phone conversations. Network users will also benefit from the new line's incredible bandwidth. Verizon said that individual customers will be able to transfer data at a blistering speed of 10-gigabits per second or higher.

Called the Trans-Pacific Express (TPE), initial capacity will be roughly 1.28-terabits per second, with a designed capacity that's upgradeable to 5.12-terabits. Verizon's vice president of operations and technology Fred Briggs said "our leadership in this project builds on our important existing relationships in China, further recognizes the emergence of China as a diverse communications hub for Asia, and reflects our company's commitment to help U.S. and other global companies compete worldwide."

According to the original press release: The cable will have a landing point provided by Verizon Business at Nedonna Beach, Ore., on the U.S. West Coast and will land on the China mainland at Qingdao and Chongming. TPE will also have landings in Tanshui, Taiwan, and Keoje, South Korea.

The project is headed up mainly by Verizon which is investing roughly $500-million USD into the project. Verizon's partners include China Telecom, China Netcom, China Unicom and several other companies in Korea and Taiwan. Surrounding countries will also benefit from the new fiber line. The TPE is slated to start construction in roughly three months and is expected to be completed by the third quarter of 2008

Google and the Third Age of Computing

To paraphrase an old comment about IBM, made during its 30 year dominance of the enterprise mainframe market, Google is not your competition, Google is the environment. Online businesses which struggle against this new reality will pay opportunity costs both in online advertising revenue as well as product success.

Health Care Blog

POLICY: Why Is Fixing American Health Care So Difficult?

Almost all the problems with the American health care system boil down to two questions. How do we create a system that ensures that all citizens, and perhaps residents, have access to health insurance? And how do we contain the huge cost increases?

Of course, behind these questions lies the question of how to reform the nation's largest industry that serves and richly rewards many powerful interests. Continue.

POLICY: Health insurance 101

PHARMA: Drug Prices here and there

There's no question that in the US big pharma and its employees colleagues at USTR are on the attack over "low drug prices" in other countries, with the assumption that the US consumer is subsidizing R&D here and therefore allowing foreigners to get the benefit of new drugs without paying their fair share. Late last year I spent some futile time (aided by Derek Lowe who writes the excellent In the Pipeline blog) trying to get at the issue of whether this is true, and whether drug profit margins abroad are so low that they wouldn't support said R&D. Of course the alternative is that drug prices are too high in the US (or at least higher than they need to be to attract investment in pharma R&D). It's a common estimate that pharma companies make about 60-70% of their profits in the US. It's also well known that net margins in the pharma industry are the highest of any major sector at roughly 17%, compared to the next highest, financial services at 14% and way above any other manufacturing industry, including software.

You don't see software companies complaining about their need for government to ensure high prices to promote R&D, although monopolies by technology standards work equally well for one well known company. Pharma companies will argue the they only have limited time to make money, as patent expiration takes away their monopoly position and essentially forces them to start again from scratch, but in any other business, competitors will also come into the market using lower prices as their wedge in. (And to be fair to big Pharma they are by no means alone in looking to regulation or tariffs to protect them--which is why American car companies concentrate on minivans and SUVs, as there's a 25% tax on foreign "light trucks"). Pharma companies also must essentially give away the fruit of their research to generic companies. So the whole issue is very complex.

PHARMA: Drug Prices here and there

Cellphone giants unveil mobile Linux foundation

Six of the world's largest telecommunications companies have officially launched a non-profit organization dedicated to creating a common mobile Linux software platform. Founded by Motorola, NEC, NTT DoCoMo, Panasonic Mobile Communications, Samsung Electronics, and Vodafone, the LiMo Foundation is inviting membership and participation from application and middleware developers.

Saturday, January 27, 2007

Exchange replacement

First thing first, why in the hell are you running KDE on a server, and more important, why are you running an X server on one at all?

A huge number of people that got stuck with Exchange servers want to get rid of them. That's why these articles keep coming up.

What you meant was that you need the address book and directory services. Scheduling tends to be done by secretaries, and forms/IMAP folder sharing is generally not needed. Now if you say you *want* scheduling, etc, then fine, there are a number of quality products from which you can choose. If you define "what you need" to be the exact feature set of Exchange, then it isn't surprising that you think you need it. You can implement everything that Exchange/Outlook does with other software, cheaper, with more reliability, and on less hardware.

1. As for AD management software... let's see. You bought Windows Server because it's easy to use and admin, Exchange because it's easy to admin, and are using AD because it's easy to admin. So to do it right, you have to buy third party software? Sounds more like somebody screwed up their research and choose a bad solution based on broken assumptions. You have to do basically the same thing on any platform, so that's not a good reason to choose one over another. The UNIX solutions are much more reliable than Exchange, too, and less expensive. They also provide all the same functionality. Unless you go out of your way to ignore the solutions that work, anyway.

2. That's because Windows' does not provide functionality such as LVM. An application can also lock a file and prevent any app with any access level from even reading it. Exchange also keeps quite a lot open and locked when it doesn't need to. If the app was written well, it wouldn't be a problem. However, your backup explaination is an excellent example of why Windows is a huge pain in the ass.

3. BS, that is a perfectly valid comparison; backing up email is backing up email. If the application is written properly, the database will be fine. Exchange isn't written well, so it has problems. That software doesn't even provide a way to do a backup without either getting third party software or shutting Exchange down. Also, your VSS stuff is essentially the *exact same thing* as LVM snapshots. Why would your way work when LVM wouldn't? If the database is inconsistent, then it's inconsistent either way.

So what you're saying is that Windows/Exchange is better because it requires more jumping through hoops, buying more random software, and more dealing with random BS like bad data formats and bad storage techniques?

Friday, January 26, 2007

The best investment advice you'll never get

Solli took one look at my unkempt collection of mut­ual funds and said, “You’re being robbed here.” He pointed to funds I had purchased from or through Putnam, Merrill Lynch, Dreyfus, and—yes—Charles Schwab (which referred me to Aperio) and asked, “Do you know that you’re paying these guys to do essentially nothing?” He carefully explained the many ingenious ways fund managers, brokers, and advisers had found to chip away at investors’ returns. Turns out that I, like more than 90 million other suckers who have put close to $9 trillion into mutual funds, was paying annual fees, commissions, and transaction costs well in excess of 2 percent a year on most of my mutual funds (see “What Are the Fees?” page 75). “Do you know what that adds up to?” Solli asked. “At the end of every 36 years, you will only have made half of what you could have, through no fault of your own. And these are fees you needn’t pay, and won’t, if you switch to index funds.”

If Solli is an industry gadfly, Geddes, a modest, unassuming son of a United Church of Christ minister, is its chainsaw massacrer. “We work in the most overcompensated industry in the country,” Geddes admitted before the water was served, “and indexing threatens the revenue flow from managed funds to brokerage houses. That’s why you’ve been kept in the dark about it. This truly is the great secret shame of our business.

Saturday, January 13, 2007

Most people who study the poor ...

'Most people who study the poor are only interested in watching them make mistakes'

How did you go around winning the confidence of the people there?

Most reporters and academics spend a few weeks or, if they are lucky, a few months, getting information and data. I spent over a decade. Many people who write about these communities are armchair critics full of false assumptions. They are not ready to get out of the hallowed halls of Ivy League institutions.

Some of them assumed that people in these depressed areas had no family values, no respect for law, and no higher aspirations. But I would soon come to question those assumptions. And because I went directly into the neglected communities with an open mind, I was able to discover many things.

My visits not only enabled people to gain confidence in me, but it also helped me to see many sides of my informants. I was able to see them grow over time. That I stayed around to watch them make mistakes and pick themselves up engendered their confidence in me -- most people who study the poor are only interested in watching them make mistakes.

Sudhir Alladi Venkatesh

Wednesday, January 10, 2007

Street Fight

Street Fight chronicles the bare-knuckles race for Mayor of Newark, N.J. between Cory Booker, a 32-year-old Rhodes Scholar/Yale Law School grad, and Sharpe James, the four-term incumbent and undisputed champion of New Jersey politics.

Fought in Newark's neighborhoods and housing projects, the battle pits Booker against an old style political machine that uses any means necessary to crush its opponents: city workers who do not support the mayor are demoted; "disloyal" businesses are targeted by code enforcement; a campaigner is detained and accused of terrorism; and disks of voter data are burglarized in the night.

Even the filmmaker is dragged into the slugfest, and by election day, the climate becomes so heated that the Federal government is forced to send in observers to watch for cheating and violence.

The battle sheds light on important American questions about democracy, power and -- in a surprising twist -- race. Both Booker and James are African-American Democrats, but when the mayor accuses the Ivy League educated Booker of not being "really black" it forces voters to examine both how we define race in this country. "We tell our children to get educated," one Newarker says, "and when they do, we call them white. What kind of a message does that send?"


Newark is a city of limitless strength and unbounded potential. It is a city of hope and promise. Newark is a city of accomplishment and a city of struggle. Newark's people, like the bricks from our Brick City nickname, are strong, resilient, enduring, and when we come together there is nothing we cannot create or achieve.

We are a people committed to strong, safe neighborhoods, to thriving schools and abundant economic opportunity. We are a city that rejoices and luxuriates in the arts and culture and have a history of setting the very rhythm for an entire nation. We are a city that boasts of great universities, houses of worship that span back decades and centuries, parks designed by America's greatest urban landscape designers and a transportation system that literally connects the globe.

Newark is a great American city. It is representative of America's success and her continuing efforts to make real the promise of this nation to everyone. Within Newark's borders lie many of America's greatest stories, some of her most courageous activists and some of her most sobering challenges. The future of our country will be defined here --by the children in our schools, the young men and women entering the workforce and by how our community meets the challenges before us. The future is now.

Street Fight tells a gripping story of the underbelly of democracy where elections are not about spin-doctors, media consultants, or photo ops. In Newark, we discover, elections are won and lost in the streets.

Tuesday, January 09, 2007

How spicy foods can kill cancers

Something that I've always known intuitively.. Thanks to Sridhar for pointing this.

Scientists have discovered the key to the ability of spicy foods to kill cancer cells.

They found capsaicin, an ingredient of jalapeno peppers, triggers cancer cell death by attacking mitochondria - the cells' energy-generating boiler rooms.

The research raises the possibility that other cancer drugs could be developed to target mitochondria.

The Nottingham University study features in Biochemical and Biophysical Research Communications.

The study showed that the family of molecules to which capsaicin belongs, the vanilloids, bind to proteins in the cancer cell mitochondria to trigger apoptosis, or cell death, without harming surrounding healthy cells.

I think it is also anti-genotoxic.

Saturday, January 06, 2007

Vegetarian diet

Assuming humans can survive on a vegetarian diet, the best argument for doing so isn't what is ethical or what we were designed for. Every time you move up a link in the food chain, only 10% of the energy of the prey is absorbed by the predator (the rest is lost as heat). So, a cow gets 10% of the grass' energy, and by eating the cow we get only 1% of the grass' energy. Eating a bird who eats bugs who eat plants, we're only getting .1% of the plant energy. Vegetarianism would be an easy way to create an abundance of food for the entire population.

I would appreciate it if you mentioned this, because this argument is rarely brought up (and I know no one else will read this comment). Thanks for the entertaining blog.

No study will ever prove or disprove the meat contention. The random, unaccountable differences between individual humans make strict science and human nutrition mutually exclusive fields.

In order for a study to say anything conclusive, the scientists would have to do something like grow 100 genetically identical clones, and then make sure each one got exactly equal amounts and types of water, sleep, stress, excercise, exposure to illness etc etc etc. Feeding half of this group a carefully controlled omnivorous diet and the other half a carefully controlled vegetarian diet with identical levels of macro and micronutrients for 12 months and then measuring their resultant "health" would be somewhere APPROACHING scientific.

Present nutritional studies typically do not have this level of rigour. Now my declatation of bias: I really like meat. Therefore I eat it.

For all the discussion on teeth, I saw a lecture on skull bones once.

From it, I learned that human teeth are:
-Not like dogs - some of their teeth are for gripping, some for chopping up bones to get at the marrow, and some for ripping flesh
-Not like cats, large or small - their teeth are mostly for ripping and slicing flesh
-Not exactly like baboons - males have huge canines for fighting other males
-Not like horses - front teeth for tearing and back teeth for grinding with a gap in between
-Not like rodents - front teeth for gnawing, always growing

-A lot like bears and pigs - people have mistaken their skulls for the skulls archaic skeletons because of the similarities between the teeth

Bears and pigs eat a lot of the same things that humans like to eat, basically everything. In fact, a certain grizzly bear in Yellowstone really liked an entire case of beer he found in a camp and was nicely drunk at the camp when they found him. But then again, when I think of bears eating meat, I think of them clawing at salmon...

Desi sites - ICICI Bank Woes

Most Indian websites are designed by retards of the highest order. In fact, each AND every site is designed by retards. The GUI is the most user-unfriendly, most things just don't work, too many ads, pop-ups. In fact, one can go on and on ...

In this series, I will attempt to cover some of them that I encounter.

Lets take for example, ICICIBank.

1. Those idiots don't let us choose our own username. They send it via mail, a week or so after we visit a bank and open an account.

2. Bank to bank transfer fails for most everyone. I just tried to transfer, resulting in several errors for reasons unknown to me. Finally, I tried to request a reissue of Transaction Password, as the transaction password expires after x number of unsuccessful attempts. Which resulted in another error..

Request for reissue of Transaction Password.


Currently, we are unable to process your request.


Please try after some time.

Tuesday, January 02, 2007

What’s Making Us Sick Is an Epidemic of Diagnoses

For most Americans, the biggest health threat is not avian flu, West Nile or mad cow disease. It’s our health-care system.

Harry Campbell

You might think this is because doctors make mistakes (we do make mistakes). But you can’t be a victim of medical error if you are not in the system. The larger threat posed by American medicine is that more and more of us are being drawn into the system not because of an epidemic of disease, but because of an epidemic of diagnoses.

Americans live longer than ever, yet more of us are told we are sick.

How can this be? One reason is that we devote more resources to medical care than any other country. Some of this investment is productive, curing disease and alleviating suffering. But it also leads to more diagnoses, a trend that has become an epidemic.

This epidemic is a threat to your health. It has two distinct sources. One is the medicalization of everyday life. Most of us experience physical or emotional sensations we don’t like, and in the past, this was considered a part of life. Increasingly, however, such sensations are considered symptoms of disease. Everyday experiences like insomnia, sadness, twitchy legs and impaired sex drive now become diagnoses: sleep disorder, depression, restless leg syndrome and sexual dysfunction.

Perhaps most worrisome is the medicalization of childhood. If children cough after exercising, they have asthma; if they have trouble reading, they are dyslexic; if they are unhappy, they are depressed; and if they alternate between unhappiness and liveliness, they have bipolar disorder. While these diagnoses may benefit the few with severe symptoms, one has to wonder about the effect on the many whose symptoms are mild, intermittent or transient.

The other source is the drive to find disease early. While diagnoses used to be reserved for serious illness, we now diagnose illness in people who have no symptoms at all, those with so-called predisease or those “at risk.”

Two developments accelerate this process. First, advanced technology allows doctors to look really hard for things to be wrong. We can detect trace molecules in the blood. We can direct fiber-optic devices into every orifice. And CT scans, ultrasounds, M.R.I. and PET scans let doctors define subtle structural defects deep inside the body. These technologies make it possible to give a diagnosis to just about everybody: arthritis in people without joint pain, stomach damage in people without heartburn and prostate cancer in over a million people who, but for testing, would have lived as long without being a cancer patient.

Second, the rules are changing. Expert panels constantly expand what constitutes disease: thresholds for diagnosing diabetes, hypertension, osteoporosis and obesity have all fallen in the last few years. The criterion for normal cholesterol has dropped multiple times. With these changes, disease can now be diagnosed in more than half the population.

Most of us assume that all this additional diagnosis can only be beneficial. And some of it is. But at the extreme, the logic of early detection is absurd. If more than half of us are sick, what does it mean to be normal? Many more of us harbor “pre-disease” than will ever get disease, and all of us are “at risk.” The medicalization of everyday life is no less problematic. Exactly what are we doing to our children when 40 percent of summer campers are on one or more chronic prescription medications?

No one should take the process of making people into patients lightly. There are real drawbacks. Simply labeling people as diseased can make them feel anxious and vulnerable — a particular concern in children.

But the real problem with the epidemic of diagnoses is that it leads to an epidemic of treatments. Not all treatments have important benefits, but almost all can have harms. Sometimes the harms are known, but often the harms of new therapies take years to emerge — after many have been exposed. For the severely ill, these harms generally pale relative to the potential benefits. But for those experiencing mild symptoms, the harms become much more relevant. And for the many labeled as having predisease or as being “at risk” but destined to remain healthy, treatment can only cause harm.

The epidemic of diagnoses has many causes. More diagnoses mean more money for drug manufacturers, hospitals, physicians and disease advocacy groups. Researchers, and even the disease-based organization of the National Institutes of Health, secure their stature (and financing) by promoting the detection of “their” disease. Medico-legal concerns also drive the epidemic. While failing to make a diagnosis can result in lawsuits, there are no corresponding penalties for overdiagnosis. Thus, the path of least resistance for clinicians is to diagnose liberally — even when we wonder if doing so really helps our patients.

As more of us are being told we are sick, fewer of us are being told we are well. People need to think hard about the benefits and risks of increased diagnosis: the fundamental question they face is whether or not to become a patient. And doctors need to remember the value of reassuring people that they are not sick. Perhaps someone should start monitoring a new health metric: the proportion of the population not requiring medical care. And the National Institutes of Health could propose a new goal for medical researchers: reduce the need for medical services, not increase it.


Should I Be Tested for Cancer?: Maybe Not

A review..

H. Gilbert Welch, MD, MPH, has written an unusually understandable revelation of the folly of testing for cancer in people with no symptoms. He explains how only a few people will benefit from common tests such as PSA, fecal blood, mammograms and others. He is enough of an insider to be able to explain the flaws in clinical trials being used by "authorities" to recommend extensive testing, and the lack of trials in some cases. The unneccessary biopsies, surgeries, radiations, chemotherapies for slow-growing cancers or even non-malignant ones are presented bravely. The uncertainty of testing is exposed where a positive for cancer may be wrong 1/3 of the time. And it is up to the patient to get second opinions.

The financial and legal pressures on MDs to test excessively are brought out. There is advice on talking or writing to your MD to indicate your unwillingness to undergo too many tests, and not to hold your MD liable if a cancer was "missed" - that is the big thing.

The deaths caused by cancer treatment are aired. This is something very few people, even MDs, know. Even when a treatment can cut the deaths from a particular cancer in half, most current treatments create non-cancer deaths, many of which will be improperly reported.

Welch is a special expert on the misleading nature of 5-year survival rates how they can rise because of early detection, yet with no change in the cancer plus cancer treatment mortality rate.

There are good explanations of how 5-year survival rates are calculated, how age-adjustments are made, how randomization for trials is done, and other things not even taught in medical school, but reserved for medical researchers. And quite easy to comprehend with clear figures and tables.

No errors that I can find; a really excellent book.