The backup problem – sometimes the backup isn’t worth the cost

Halamka has a great review of backup strategies and costs, but my favorite bit is in the last paragraph …

Life as a Healthcare CIO: Our Storage Backup Strategy

Over the past year, Harvard Medical School has worked with research, administrative, and educational stakeholders to develop a set of storage policies and technologies that support demand, are achievable in the short term and are affordable.

I recently gave a keynote at Bio-IT World where I described the HMS storage strategy to ensure scalability, high performance, and reliability.

Since that presentation, we've refined our strategy for replication/backup/restoration of data for disaster recovery. In many ways backup is a harder problem to solve and a more expensive project than data storage itself.

Our best thinking (a strawman for now that we are still reviewing with customers) is outlined on this slide. For databases and Microsoft exchange, we're using Data Domain appliances to replace tape …

…  Some departments have asked not to replicate at all, since it is cheaper to rerun an experiment than to replicate the terabytes of data each experiment generates. …

I recently sat through a fascinating recounting of a corporate IT outage. They thought they had sufficient redundancy, but there’s always a limit.

Backups aren’t just a problem for home users. Our current technologies don’t scale as well as one might imagine.

Gmail: I don’t love you any more, but we can still be friends

I’ve been using Gmail since the week it was “released”. There’s a lot I like about it, but I’ve finally decided it’s killing me.

There’s more than one set of problems, but the number one problem is the bloody obligate subject line threading model.

It wouldn’t be so bad if they used some kind of message unique identifier to implement usenet style threading. It would be immensely better if Gmail allowed me to edit the subject lines of incoming messages and thus to create new threads (like this).

In the current state though, it’s a killer. I have too much trouble picking out critically important messages from threads. Searches return the thread, and trawling through the thread is too error prone.

The obligate threading was and is a mistake.

There are other Gmail problems. Google’s only recently fixed up the Contacts model, and I despise the UI for creating mailing lists and working with Contacts. Gmail can be intermittently slow and unreliable – this past week has been very annoying. The IMAP implementation intersects badly with labels; my use of labels means Spotlight searches in OS X Mail.app return multiple instances of a single message.

There’s still a lot I like about Gmail, especially when I use it with iPhone Mail.app. I’m not going to abandon it, but I’m considering simplification measures. I might return to POP style access on OS X for example and use OS X Mail.app when I need to get real work done.

As long as Gmail was used primarily for personal email, their lowest common denominator approach wasn’t necessarily wrong. That’s often a good way to win. Now, though, Google wants to support large enterprises on their Google Apps platform. There are going to be more users like me.

If Google doesn’t start listening to its more demanding customers I won’t be the only one to start seeing other software.

Update: I realize that Gmail contributes to hiding messages in mis-identified threads by hiding the subject line on reply. Sigh.

Choosing a brand? check GoodGuide

I'm delighted someone's doing this ...

An App, the GoodGuide, Aids in Careful Shopping - NYTimes.com

... GoodGuide, a Web site and iPhone application that lets consumers dig past the package’s marketing spiel by entering a product’s name and discovering its health, environmental and social impacts....

Kellogg got an OK rating.

Subversive theophysics - Greg Egan

I've been composing a post about Greg Egan's Permutation City for a while. I'm afraid I'll never get to the whole thing, so I'm going to toss off the short version. (Warning, contains spoilers)

Greg Egan is usually said to write "hard" science fiction. That's inadequate. He writes neutronium grade science fiction. His mathematical physics bent has become so extreme that his latest book is a thin layer of fiction around a core of speculative physics (Amazon promises me a copy in 3-4 weeks, apparently they have to retype it. Egan has put a prequel to the story on his web site).

Permutation City is one of his best works. Despite the math science bent several of the characters have stuck with me.

The best part though, is the fusion between theology and physics -- theophysics. In Permutation City reality is fundamentally mathematical, much as imagined by Stephen Wolfram and many more conventional physicists. A group of experimental modelers creates an artificial world with a different sort of mathematical reality.

No wait, hang in here for a minute. I'm really going somewhere.

The creatures of this new world are fantastically alien, but like us they're compelled to understand their world. Problem is, their world is fundamentally incomprehensible. It was created by omniscient and omnipotent Creators. Gods.

So the alien critter(s) is(are) "anguished". They are compelled to understand, but they cannot understand. The human Creators are sympathetic, and decide to manifest themselves in the alien world. The Truth shall be known, and the aliens will understand.

Except, the aliens come up with their own Theory of Everything; their equivalent of quantum gravity. It looks crazy and absurd, but it's internally consistent. It explains everything but the appearance of the Creators, and that detail can be quickly forgotten.

The Creators suddenly find themselves written out of the script, but that's a different story. I'm telling the story of the subversive aspects of Egan's fiction.

Obviously, the invented aliens of Permutation City aren't alone. We too are compelled to comprehend, and modern physics is getting pretty damned absurd...

State of the Art; The Origin of Life

This article does a great job of describing one of biology's great questions, and describing the state of the art. Superb work. Unfortunately it's so good I'm obliged, for the sake of my own reference, to replicate the whole darned thing. Please click on the link so the NYT gets a visit. Emphases mine.

New Glimpses of Life’s Puzzling Origins - Nicholas Wade - NYTimes.com

.... 3.9 billion years ago, a shift in the orbit of the Sun’s outer planets sent a surge of large comets and asteroids careening into the inner solar system. Their violent impacts gouged out the large craters still visible on the Moon’s face, heated Earth’s surface into molten rock and boiled off its oceans into an incandescent mist.

Yet rocks that formed on Earth 3.8 billion years ago, almost as soon as the bombardment had stopped, contain possible evidence of biological processes. If life can arise from inorganic matter so quickly and easily, why is it not abundant in the solar system and beyond? If biology is an inherent property of matter, why have chemists so far been unable to reconstruct life, or anything close to it, in the laboratory?

The origins of life on Earth bristle with puzzle and paradox. Which came first, the proteins of living cells or the genetic information that makes them? How could the metabolism of living things get started without an enclosing membrane to keep all the necessary chemicals together? But if life started inside a cell membrane, how did the necessary nutrients get in?

The questions may seem moot, since life did start somehow. But for the small group of researchers who insist on learning exactly how it started, frustration has abounded. Many once-promising leads have led only to years of wasted effort. Scientists as eminent as Francis Crick, the chief theorist of molecular biology, have quietly suggested that life may have formed elsewhere before seeding the planet, so hard does it seem to find a plausible explanation for its emergence on Earth.

In the last few years, however, four surprising advances have renewed confidence that a terrestrial explanation for life’s origins will eventually emerge.

One is a series of discoveries about the cell-like structures that could have formed naturally from fatty chemicals likely to have been present on the primitive Earth. This lead emerged from a long argument between three colleagues as to whether a genetic system or a cell membrane came first in the development of life. They eventually agreed that genetics and membranes had to have evolved together.

The three researchers, Jack W. Szostak, David P. Bartel and P. Luigi Luisi, published a somewhat adventurous manifesto in Nature in 2001, declaring that the way to make a synthetic cell was to get a protocell and a genetic molecule to grow and divide in parallel, with the molecules being encapsulated in the cell. If the molecules gave the cell a survival advantage over other cells, the outcome would be “a sustainable, autonomously replicating system, capable of Darwinian evolution,” they wrote.

“It would be truly alive,” they added.

One of the authors, Dr. Szostak, of the Massachusetts General Hospital, has since managed to achieve a surprising amount of this program.

Simple fatty acids, of the sort likely to have been around on the primitive Earth, will spontaneously form double-layered spheres, much like the double-layered membrane of today’s living cells. These protocells will incorporate new fatty acids fed into the water, and eventually divide.

Living cells are generally impermeable and have elaborate mechanisms for admitting only the nutrients they need. But Dr. Szostak and his colleagues have shown that small molecules can easily enter the protocells. If they combine into larger molecules, however, they cannot get out, just the arrangement a primitive cell would need. If a protocell is made to encapsulate a short piece of DNA and is then fed with nucleotides, the building blocks of DNA, the nucleotides will spontaneously enter the cell and link into another DNA molecule.

At a symposium on evolution at the Cold Spring Harbor Laboratory on Long Island last month, Dr. Szostak said he was “optimistic about getting a chemical replication system going” inside a protocell. He then hopes to integrate a replicating nucleic acid system with dividing protocells.

Dr. Szostak’s experiments have come close to creating a spontaneously dividing cell from chemicals assumed to have existed on the primitive Earth. But some of his ingredients, like the nucleotide building blocks of nucleic acids, are quite complex. Prebiotic chemists, who study the prelife chemistry of the primitive Earth, have long been close to despair over how nucleotides could ever have arisen spontaneously.

Nucleotides consist of a sugar molecule, like ribose or deoxyribose, joined to a base at one end and a phosphate group at the other. Prebiotic chemists discovered with delight that bases like adenine will easily form from simple chemicals like hydrogen cyanide. But years of disappointment followed when the adenine proved incapable of linking naturally to the ribose.

Last month, John Sutherland, a chemist at the University of Manchester in England, reported in Nature his discovery of a quite unexpected route for synthesizing nucleotides from prebiotic chemicals. Instead of making the base and sugar separately from chemicals likely to have existed on the primitive Earth, Dr. Sutherland showed how under the right conditions the base and sugar could be built up as a single unit, and so did not need to be linked.

“I think the Sutherland paper has been the biggest advance in the last five years in terms of prebiotic chemistry,” said Gerald F. Joyce, an expert on the origins of life at the Scripps Research Institute in La Jolla, Calif.

Once a self-replicating system develops from chemicals, this is the beginning of genetic history, since each molecule carries the imprint of its ancestor. Dr. Crick, who was interested in the chemistry that preceded replication, once observed, “After this point, the rest is just history.”

Dr. Joyce has been studying the possible beginning of history by developing RNA molecules with the capacity for replication. RNA, a close cousin of DNA, almost certainly preceded it as the genetic molecule of living cells. Besides carrying information, RNA can also act as an enzyme to promote chemical reactions. Dr. Joyce reported in Science earlier this year that he had developed two RNA molecules that can promote each other’s synthesis from the four kinds of RNA nucleotides.

“We finally have a molecule that’s immortal,” he said, meaning one whose information can be passed on indefinitely. The system is not alive, he says, but performs central functions of life like replication and adapting to new conditions.

“Gerry Joyce is getting ever closer to showing you can have self-replication of RNA species,” Dr. Sutherland said. “So only a pessimist wouldn’t allow him success in a few years.”

Another striking advance has come from new studies of the handedness of molecules. Some chemicals, like the amino acids of which proteins are made, exist in two mirror-image forms, much like the left and right hand. In most naturally occurring conditions they are found in roughly equal mixtures of the two forms. But in a living cell all amino acids are left-handed, and all sugars and nucleotides are right-handed.

Prebiotic chemists have long been at a loss to explain how the first living systems could have extracted just one kind of the handed chemicals from the mixtures on the early Earth. Left-handed nucleotides are a poison because they prevent right-handed nucleotides linking up in a chain to form nucleic acids like RNA or DNA. Dr. Joyce refers to the problem as “original syn,” referring to the chemist’s terms syn and anti for the structures in the handed forms.

The chemists have now been granted an unexpected absolution from their original syn problem. Researchers like Donna Blackmond of Imperial College London have discovered that a mixture of left-handed and right-handed molecules can be converted to just one form by cycles of freezing and melting.

With these four recent advances — Dr. Szostak’s protocells, self-replicating RNA, the natural synthesis of nucleotides, and an explanation for handedness — those who study the origin of life have much to be pleased about, despite the distance yet to go. “At some point some of these threads will start joining together,” Dr. Sutherland said. “I think all of us are far more optimistic now than we were five or 10 years ago.”

One measure of the difficulties ahead, however, is that so far there is little agreement on the kind of environment in which life originated. Some chemists, like Günther Wächtershäuser, argue that life began in volcanic conditions, like those of the deep sea vents. These have the gases and metallic catalysts in which, he argues, the first metabolic processes were likely to have arisen.

But many biologists believe that in the oceans, the necessary constituents of life would always be too diluted. They favor a warm freshwater pond for the origin of life, as did Darwin, where cycles of wetting and evaporation around the edges could produce useful concentrations and chemical processes.

No one knows for sure when life began. The oldest generally accepted evidence for living cells are fossil bacteria 1.9 billion years old from the Gunflint Formation of Ontario. But rocks from two sites in Greenland, containing an unusual mix of carbon isotopes that could be evidence of biological processes, are 3.830 billion years old.

How could life have gotten off to such a quick start, given that the surface of the Earth was probably sterilized by the Late Heavy Bombardment, the rain of gigantic comets and asteroids that pelted the Earth and Moon around 3.9 billion years ago? Stephen Mojzsis, a geologist at the University of Colorado who analyzed one of the Greenland sites, argued in Nature last month that the Late Heavy Bombardment would not have killed everything, as is generally believed. In his view, life could have started much earlier and survived the bombardment in deep sea environments.

Recent evidence from very ancient rocks known as zircons suggests that stable oceans and continental crust had emerged as long as 4.404 billion years ago, a mere 150 million years after the Earth’s formation. So life might have had half a billion years to get started before the cataclysmic bombardment.

But geologists dispute whether the Greenland rocks really offer signs of biological processes, and geochemists have often revised their estimates of the composition of the primitive atmosphere. Leslie Orgel, a pioneer of prebiotic chemistry, used to say, “Just wait a few years, and conditions on the primitive Earth will change again,” said Dr. Joyce, a former student of his....

A cogent, and funny, observation on health care costs

I don't agree with the premise, but I rather liked Gail Collins' aside ...

Health Care Follies - The Conversation Blog - Gail Collins - NYTimes.com

... The big problem is that the economy is sinking under the rising cost of medical treatment, one cause of which is doctors recommending unnecessary and overly costly procedures. (For which I do not blame the doctors. If we lived in a desirable world in which people were insured against not getting enough news, I can guarantee you that I would come up with some really excellent additional products.)...

She really is a wonderful writer, and she's getting better.

Unfortunately, as recently illustrated by Atul Gawande, the premise is sadly simplistic. It's not simply that all these procedures are "unnecessary", it's rather that most people, if they had to pay for them, would choose something cheaper even it were less effective.

So if we had to pay for our shoulder MRIs, we'd probably give "rest" a longer try before looking for a (probably inoperable) rotator cuff tear. If we had to pay for our anterior cruciate repair, we might decide to live without inline skating -- like we used to.

This is why these decisions are much more troublesome than anyone but an unpaid blogger is willing to publicly acknowledge.

We won't really get universal coverage until we close our eyes, grit our teeth, seal our nostrils, and embrace crummy care (aka "good enough" care).

Health War II

America's Health War I ended about 14 years ago.

We lost, the bad guys won.

Now Health War II has begun ...

Roert Reich's Blog: The Healthcare War is Now Official

Yesterday the American Medical Association came out against a public option for health care. And yesterday the President reaffirmed his support for it...

... All major lobbying firms in Washington -- many of them brimming with ex-members of Congress -- are now crawling all over the Hill. Lots of money is on the table. AMA's political action committee has contributed $9.8 million to congressional candidates since 2000, and its lobbying arm is one of the most formidable on the Hill. Meanwhile, Big Insurance and Big Pharma are increasing their firepower. The five largest private insurers and their trade group America's Health Insurance Plans spent a total of $6.4 million on lobbying in the first quarter of this year, up more than $1 million from the first quarter last year, and are spending even more now. United Health Group spent $1.5 million in the first quarter, up 34 percent from the $1.1 million it spent in the first quarter last year. Aetna spent $809,793 between January and the end of March, up 41 percent from last year. Pfizer, the world's biggest drugmaker, spent more than $6.1 million on lobbying between January and March, more than double what it spent last year. It also spent nearly $3.3 million lobbying in the fourth quarter of 2008. Every one of them is upping their spending....

... The President can't do this alone. You must weigh in and get everyone you know to weigh in, too. Bombard your senators and representatives. Organize and mobilize others. And let the White House know how strongly you feel...

In Minnesota our senators are Al Franken (still not seated) and Amy Klobuchar. I'm not worried about Franken, he'll do the right thing. I'm very concerned about Klobuchar. I've sent her an email through her extremely sluggish contact form.

This won't be pretty. I'd expect Obama to lose this one, except he's a pretty resourceful guy.

We know where we have to end up. We need to get to "Crummy Care".

Crummy Care is healthcare with dingy carpets and peeling (lead free) paint. Crummy Care is delivered by physicians, but also by a lot of cheaper routes. Crummy Care sends films to be (double) read in Malaysia, uses CT scanners that are four years past cutting edge, has waiting lists for services, and has tight formularies and highly optimized low cost standard care protocols.

Crummy Care is not what most of us want, but it is what we can afford to provide to everyone in America. (No, I'm sorry, electronic health records aren't really a miracle solution -- though they can make Crummy Care quite a bit better.)

Crummy Care is the good enough, affordable, minimalist solution. If you have money you can buy Mercedes care, but if you don't you'll get reasonably state of the art treatment for your breast cancer (sorry, probably not a bone marrow transplant) and reasonable preventive care (no PSA for your prostate).

So we know where we have to go. The problem is it's a lousy, harsh, innocence crushing trip. We won't travel that road if we try do do business as usual

Which is why we need the "public option".

If we don't get it, then I think we'll eventually decide we lost Health War II.

So what happens if we lose?

We wait 5-7 years for Health War III.

The Dept of Energy joins the peak oil parade

They're weaving and dodging, but basically they're on board ...

Goodbye to cheap oil | Salon

Every summer, the Energy Information Administration of the U.S. Department of Energy issues its International Energy Outlook (IEO) — a jam-packed compendium of data and analysis on the evolving world energy equation...

... The IEO predicts a sharp drop in projected future world oil output (compared with previous expectations) and a corresponding increase in reliance on what are called "unconventional fuels" — oil sands, ultra-deep oil, shale oil and biofuels...

... 81.5 million barrels produced in 2006 ... projected 2030 figure to just 93.1 million barrels per day...

They're much more pessimistic than they were about three years ago. They're not saying we'll run out of hydrocarbons, in fact they predict our 2030 oil production will be about 13% higher than it was 3 years ago. The catch is worldwide energy demand would normally rise by far more than 13% over the next 20 years.

That demand won't be filled by traditional oil supplies. We'll either do more with far less energy or we'll bake the planet with CO2 producing oil alternatives.

Peak oil is upon us. Expect your gas price to continue to rise.

Google's problem: too many Starters, too few Finishers

Google's Blogger rich text editor is richly broken in Firefox and Safari. This isn't a new problem, it's been broken in lesser ways for years.  You might image you can fix the broken formatting in the "Edit HTML" view, but that view is essentially obsolete. It doesn't provide full access to the formatting controls.

Google's My Maps is stuck in toy stage.

There are lots of other examples. Google is great at starting products, lousy at finishing them.

This isn't surprising. Google has a famously IQ and creativity focused recruiting policy, a policy strongly identified with Marissa Mayer.

That's a great way to get Starters. Starters are people who invent things.

It's not a great way to get Finishers. Finishers are the people who do the 80% of the work required for the last 20% of the product. Finishers are detail people.

Google's problem is too many Starters, too few Finishers. Google needs to change their recruiting and hiring policies.

Bubble III

Governments have injected trillions of dollars into financial systems. We're not set up to quickly deploy that much money, and loans are still slow. The big infrastructure projects are still months to years away.

The money has to go somewhere though. So where's it going? The market.

A few months ago, in a fit of contrarian optimism, I predicted Dow 9000 by Nov 2009. It's June and the Dow is at 8800; we could easily hit 9000 this month. That's not necessarily good. A November 9000 could be compatible with a reasonable recovery, but a June 9000 is crackers.

I think I hear something inflating ...

The mass murder of Navy sonar

I don't think the US Navy would get away with deploying technology that visibly tortured and killed large numbers of land mammals ...

Does Military Sonar Kill Marine Wildlife?: Scientific American:

... generate slow-rolling sound waves topping out at around 235 decibels; the world’s loudest rock bands top out at only 130. These sound waves can travel for hundreds of miles under water, and can retain an intensity of 140 decibels as far as 300 miles from their source...

... Navy documents which estimated that such testing would kill some 170,000 marine mammals and cause permanent injury to more than 500 whales, not to mention temporary deafness for at least 8,000 others...

Imagine the pain of being unable to escape this barrage.

It’s not Obama’s deficit

No great surprise, but even I’m surprised how relatively little Obama’s budgets have contributed to the Federal deficit …

How the U.S. Surplus Became a Deficit – David Leonhardt - NYTimes.com

… The New York Times analyzed Congressional Budget Office reports going back almost a decade, with the aim of understanding how the federal government came to be far deeper in debt than it has been since the years just after World War II. This debt will constrain the country’s choices for years and could end up doing serious economic damage if foreign lenders become unwilling to finance it…

…You can think of that roughly $2 trillion swing as coming from four broad categories: the business cycle, President George W. Bush’s policies, policies from the Bush years that are scheduled to expire but that Mr. Obama has chosen to extend, and new policies proposed by Mr. Obama.

The first category — the business cycle — accounts for 37 percent of the $2 trillion swing. It’s a reflection of the fact that both the 2001 recession and the current one reduced tax revenue, required more spending on safety-net programs and changed economists’ assumptions about how much in taxes the government would collect in future years.

About 33 percent of the swing stems from new legislation signed by Mr. Bush. That legislation, like his tax cuts and the Medicare prescription drug benefit, not only continue to cost the government but have also increased interest payments on the national debt.

Mr. Obama’s main contribution to the deficit is his extension of several Bush policies, like the Iraq war and tax cuts for households making less than $250,000. Such policies — together with the Wall Street bailout, which was signed by Mr. Bush and supported by Mr. Obama — account for 20 percent of the swing.

About 7 percent comes from the stimulus bill that Mr. Obama signed in February. And only 3 percent comes from Mr. Obama’s agenda on health care, education, energy and other areas.

So we’re talking 37% business cycle, 33% Bush, 20% Bush items extended by Obama (tax cuts, Iraq) and 10% Obama. It’s the last 10% that makes GOP budget “hawks” rabid.

Insofar that Obama’s spending is a stimulus package aimed at reducing the duration of the recession accounting for 37% of the deficit, it’s an excellent investment.

The Pre advantage: $1,200 cheaper

The Pre has three significant advantages over iPhone 3.0 [1].

1. PIM. Palm is far more committed to productivity (PIM) apps (calendar, contacts, tasks, notes) than Apple. To a first approximation Apple despises this entire domain.
2. Sprint. AT&T leads the League of Evil with cramming, rebate scams, SMS spam, overselling network capacity, and other low class trickery. Sprint is merely wicked, but AT&T is the Microsoft of phone companies.
3. Price.

The last is a biggie ...

We know the iPhone isn't cheap, but Billshrink shows the numbers

... , the cheapest phone to own over a two-year period is the Palm Pre, which clocks in at $2,400. The cost calculated is based off of a service plan with two years of unlimited voice, data, and messaging services. The Android G1 follows with $3,240 and finally the iPhone 3G S at $3,600...

I hope these are enough advantages to either make Palm viable or get a committed buyer for the company.

[1] I think limited battery life will make Palm multitasking a modest improvement over iPhone 3 notifications.

Google's HTML 5 - the netbook chromestellation platform

It’s been a few months since I’ve last written about the netbook chromestellation platform (Google Chrome + Netscape Constellation 1996). There’s been little of interest on the netbook side, except that Microsoft has successfully executed their only logical response – make XP effectively free on the Netbook. Price points have not yet fallen into the truly disruptive range.

On the software side though, we now have Google’s technology preview from their I/O conference: Google I/O - Google's HTML 5 Work: What's Next?.

One surprise is what a large role Apple has played. Clearly they made the right strategic choice when they split from Mozilla to create Safari/WebKit and introduced Canvas. Apple and Google are precisely aligned now, and both are pushing the vision of a browser as a technology platform. (From a very different direction, so is Palm.)

Another surprise is that the standards groups have abandoned the slow pace of seven years ago and resumed the 1990s model of adopting whatever the leaders are doing. HTML “5” today is nothing like HTML “5” of 2004.

It’s everyone against Microsoft now, rather than the 1990s Sun vs. Netscape vs. Microsoft battle (Microsoft hardly had to lift a finger in that one – Netscape and Sun killed one another). If you add the EU in then maybe it’s almost a fair fight, but Microsoft is immensely profitable and Windows 7 is due*.

So now we wait for the Google branded Cloud-powered $150 HTML 5 netbook …

* Windows 7 will sell so much hardware and software, it might inflate .com bubble 2.0.

iPhone 3GS - interesting disability features

The iPhone 3GS should be able to support magnifying text and acting as a reader for printed materials.

I expect that 'magnifying glass' apps will soon be as popular as flashlight apps.

Extra points to the first to combine the flashlight and magnifier features in a single app.