Thursday, April 05, 2007

Cheney/Bush and torture - now it's sensory deprivation

I wonder what they expect to get out of the prisoners now ...
BBC NEWS | Americas | Guantanamo conditions 'worsening': "

... The report, published on Thursday, said about 300 detainees are now being held at a new facility - known as Camp 5, Camp 6 and Camp Echo - comparable to 'super-max' high security units in the US.

The group said the facility had 'created even harsher and apparently more permanent conditions of extreme isolation and sensory deprivation'.

It said the detainees were reportedly confined to windowless cells for 22 hours a day, only allowed to exercise at night and could go for days without seeing daylight...
Smells like revenge.

Jobs predicts the death of audio DRM within a year

A cursory examination of my DRM posts would show that, until the end of 2006, I didn't expect audio DRM to fall so quickly. It appears in retrospect to have been a rotted tree, ready to collapse.

If Jobs is correct in his prediction, this means the effective end of audio DRM within the year ...
AppleInsider | Jobs talks new iTunes functions, DRM and video, iPod storage [transcript]

...Q: You mentioned that 2.5 of 5 million [songs in your iTunes] catalog will be DRM-free by the end of the year. That's presumably not just EMI Records by the end of the year.
A: That's correct. That's our estimate. That's EMI content plus other content from other labels.
Q: And the second part of my question is: What do you think this will have an impact on the iPod-iTunes relationship in terms of now being able to buy your music on any player and not just on iPod?
A: Well again, you've been able to play all sorts of music on iPod forever. iPods have played MP3s forever. So the only music that has been in question is music you buy off the iTunes store. Now again, you can burn a CD and read that CD back in and it takes off the DRM. So you could then play it on anything else. We compete based on having what we we think is the best music store and based on what we think is having the best music players. And if customers agree with us, we are going to do well. If they don't, well we're going to get a message back that we have to work harder...
Jobs tries to imply this isn't that big a deal since consumers could always burn a CD then re-encode the music. In reality anyone who's tried this will tell you the quality is similar to AM radio.

I suspect the non-DRMd files will have, embedded within them, an encrypted and non-removable identifier that can be tracked back to the purchaser's credit card records. I'm surprised nobody has asked about that. It's what I'd do if I were Apple. (Update 5/3/07: and that's what they did, but the one you can see isn't even encrypted. I'm sure there's an encrypted tag as well.)

I'm fine with the price increase, I'll upgrade the DRMd music I like. This will make me significantly more likely to use the iTunes store.

BTW, DRM on TV and video isn't dead yet. I'm not so sure it will die. What killed audio DRM for me was that it severely impaired my ability to listen to music and audio while working or driving. I can't multitask video inputs, so I am much less impacted by DRM on video. I do buy DRMd TV shows and movies for the kids to watch while traveling ...

The return of VRML? Google 3D warehouse

Remember VRML? Virtual Reality Markup Language was an SGML-like standard for describing 3 dimensional objects and their relationships. I thought of that when I read of an OS X application that uses objects from Google's 3D Warehouse.

I haven't followed the sketch-up story very closely, but the emergence of applications that draw from an open (?) worldwide library of 3 D shapes got my attention. An interesting development!

Wednesday, April 04, 2007

Christianity and data: Not promising

Dyer reads the sociology literature ...
Dyer: Religion and Good Behavior

... the numbers don't lie: the more religious a [Christian] country is, the worse people behave in their private lives. Thank God they didn't do a survey on the correlation between strong religious belief and war.
Well, of course Dyer knows numbers lie all the time. Still, there does not appear to be any empirical support for the theory that religion leads to better behavior, even when that behavior is defined in Christian terms (turn the other cheek, help the poor, etc). Dyer suspects they're finding correlation rather than causation and I tend to agree, but one cannot rule out that Christianity causes bad behavior.

Dyer states the obvious about the Iranian snatch operation

As usual, Dyer says out loud what the insiders say amongst themselves (and the regular media never mentions)....
Iran: How to Start a War:

... it was not necessarily an operation ordered from the top of Iran's government. In fact, there is no single source of authority in Iran's curious system of 'multiple governments,' as one observer labelled the impenetrably complex division of responsibilities and powers between elected civilians and unelected mullahs. The Revolutionary Guards (who are quite different from the regular armed forces) enjoy considerable autonomy within this system.

According to the US authorities in Iraq, the five Iranian diplomats arrested by US troops in a raid in Irbil in Iraqi Kurdistan last January were actually Revolutionary Guards, and it would seem that their colleagues want them back. Kidnapping American troops as hostages for an exchange could cause a war, so they decided to grab some Brits instead. And it will probably work, after a certain delay...
Dyer points out that the Iranians had a full ambush waiting for the Brits, and that it's silly to think they didn't have a logical agenda. When I heard the story I assumed it was a swap maneuver, but Dyer fills in details on why the Iranians wanted Brits and not Americans. (Yanks are more aggressive, hence a snatch would be more likely to produce dead US soldiers and an invasion of Iran -- both counter-productive.)

The Poincare Conjecture, In Our Time, and the limits of human minds

My favorite radio show (podcast) covered the The Poincare Conjecture last November. I just listened to it this week, which is the beauty of podcasting of course. I wouldn't have gotten nearly as much from a 20th century broadcast -- I lost count of how often I hit the replay button. The "lasso" analogies in particular were a bit mangled, and I think Melvyn may have skipped over what the conjecture actually is (one could decipher it by context; perhaps I failed to hit the replay button often enough).

It's a great show, but they needed another hour to do justice to Poincaré and to the lay version of his mathematics. More significantly, I think IOT hit the limits of what the medium can handle with this topic; Melvynn kept half-joking that his head was hurting. The one non-mathematician guest had to opt out of a final comment, saying "this is too hard for me". The listener comments reflect how challenging the topic was.

Which brings me to my recent readings in physics. I'm reading Gribbins on Quantum Mechanics (1994) and Brian Greene with another cosmology/string theory overview, the Gribbins book is my personal favorite, but it's a bit dated now. Together though, they make it hard to overlook that physics seems to be getting harder and harder. We have more physicists than ever, and I'd wager there's a Feynman or two in the bunch, but we've been stuck for decades now. String theory's luster is a bit tarnished (Smolin et al), special relativity doesn't fit with 'spooky action at a distance', we can't unify gravity and QM, the math keeps getting harder (or so I read) and the universe feels ever more absurd and ill-suited to the frailties of the human mind.

Sort of like Perelman's proof of the Poincaré conjecture, which teams of mathematicians are still slogging through years after its unofficial publication. Modern mathematics is said to be so complex only hyper-specialists can tackle it, but hyper-specialists can't make the kind of cross-domain breakthroughs Poincaré made.

I wonder if we'll run aground until we come up with better minds -- either human or otherwise.

PS. Firefox spell checking is very impressive. It corrected my spelling of Melvyn Bragg's first name!

Tuesday, April 03, 2007

Pet Poisoning 2007: SFGate has the best overview

The bottom line continues for me continues to be -- time to move pets to food that falls under human regulation. SF Gate has the best recent coverage (emphases mine):
YOUR WHOLE PET / Bigger than you think: The story behind the pet food recall

... I'm a contributing editor for a nationally syndicated pet feature, Universal Press Syndicate's Pet Connection, and all of us there have close ties to the veterinary profession. Two of our contributors are vets themselves, including Dr. Marty Becker, the vet on "Good Morning America." And what we were hearing from veterinarians wasn't matching what we were hearing on the news.

When we started digging into the story, it quickly became clear that the implications of the recall were much larger than they first appeared. Most critically, it turned out that the initially reported tally of dead animals only included the cats and dogs who died in Menu's test lab and not the much larger number of affected pets.

Second, the timeline of the recall raised a number of concerns. Although there have been some media reports that Menu Foods started getting complaints as early as December 2006, FDA records state the company received their first report of a food-related pet death on February 20.

One week later, on February 27, Menu started testing the suspect foods. Three days later, on March 3, the first cat in the trial died of acute kidney failure. Three days after that, Menu switched wheat gluten suppliers, and 10 days later, on March 16, recalled the 91 products that contained gluten from their previous source.

Nearly one month passed from the date Menu got its first report of a death to the date it issued the recall. During that time, no veterinarians were warned to be on the lookout for unusual numbers of kidney failure in their patients. No pet owners were warned to watch their pets for its symptoms. And thousands and thousands of pet owners kept buying those foods and giving them to their dogs and cats.

At that point, Menu had seen a 35 percent death rate in their test-lab cats, with another 45 percent suffering kidney damage. [jf: of those exposed, 80% were injured or killed] The overall death rate for animals in Menu's tests was around 20 percent. How many pets, eating those recalled foods, had died, become ill or suffered kidney damage in the time leading up to the recall and in the days since? The answer to that hasn't changed since the day the recall was issued: We don't know.

We at Pet Connection knew the 10-15 deaths being reported by the media did not reflect an accurate count. We wanted to get an idea of the real scope of the problem, so we started a database for people to report their dead or sick pets. On March 21, two days after opening the database, we had over 600 reported cases and more than 200 reported deaths. As of March 31, the number of deaths alone was at 2,797.

There are all kinds of problems with self-reported cases, and while we did correct for a couple of them, our numbers are not considered "confirmed." But USA Today reported on March 25 that data from Banfield, a nationwide chain of over 600 veterinary hospitals, "suggests [the number of cases of kidney failure] is as high as hundreds a week during the three months the food was on the market."

On March 28, "NBC News" featured California veterinarian Paul Pion, who surveyed the 30,000 members of his national Veterinary Information Network and told anchor Tom Costello, "If what veterinarians are suspecting are cases, then it's much larger than anything we've seen before." Costello commented that it amounted to "potentially thousands of sick or dead pets."

The FDA was asked about the numbers at a press conference it held on Friday morning to announce that melamine had been found in the urine and tissues of some affected animals as well as in the foods they tested. Dr. Stephen Sundlof, director of the Center for Veterinary Medicine, told reporters that the FDA couldn't confirm any cases beyond the first few, even though they had received over 8,800 additional reports, because "we have not had the luxury of confirming these reports." They would work on that, he said, after they "make sure all the product is off the shelves." He pointed out that in human medicine, the job of defining what constitutes a confirmed case would fall to the Centers for Disease Control, but there is no CDC for animals.

Instead, pet owners were encouraged to report deaths and illness to the FDA. But when they tried to file reports, there was no place on the agency's Web site to do so and nothing but endless busy signals when people tried to call.

Veterinarians didn't fare much better. They were asked to report cases to their state veterinarian's office, but one feline veterinary blog, vetcetera, which surveyed all official state veterinarian Web sites, found that only eight had any independent information about the recall, and only 24 even mentioned it at all. Only one state, Vermont, had a request on their site for veterinarians to report pets whose illnesses or deaths they suspect are related to the recall. And as of today, there is no longer a notice that veterinarians should report suspected cases to their state veterinarians on the Web site of the American Veterinary Medical Association.

The lack of any notification system was extremely hard on veterinarians, many of whom first heard about the problem on the news or from their clients. Professional groups such as the Veterinary Information Network were crucial in disseminating information about the recall to their members, but not all vets belong to VIN, and not all vets log on to VIN on the weekend (the Menu press release, like most corporate or government bad news, was issued on a Friday)...

...How did this problem, now involving almost every large pet food company in the United States, including some of the most trusted -- and expensive -- brands, get so out of hand? How come pet owners weren't informed more rapidly about the contaminated pet food? Why is it so hard to get accurate numbers of affected animals? Why didn't veterinarians get any notification? Where did the system break down?

The issue may not be that the system broke down, but that there isn't really a system.

There is, as the FDA pointed out, no veterinary version of the CDC. This meant the FDA kept confirming a number it had to have known was only the tip of the iceberg. It prevented veterinarians from having the information they needed to treat their patients and advise pet owners. It allowed the media to repeat a misleadingly low number, creating a false sense of security in pet owners -- and preventing a lot of people from really grasping the scope and implication of the problem.

... On Sunday night, April 1, Pet Connection got a report from one of its blog readers, Joy Drawdy, who said that she had found an import alert buried on the FDA Web site. That alert, issued on Friday, the same day that the FDA held its last press conference about the recall, identified the Chinese company that is the source of the contaminated gluten -- gluten that is now known to be sold not only for use in animal feed, but in human food products, too. (The Chinese company is now denying that they are responsible, although they are investigating it.)

Although the FDA said on Friday it has no reason to think the contaminated gluten found its way into the human food supply, Sundlof told reporters that it couldn't be ruled out. He also assured us that they would notify the public as soon as they had any more information -- except, of course, that they did have more information and didn't give it to us, publishing it instead as an obscure import alert, found by chance by a concerned pet owner, which was then spread to the larger media.

All of which begs the question: If a system to report and track had been in place for animal illness, would this issue have emerged sooner? Even lacking a reporting and tracking system, if the initial news reports had included, as so many human stories do, suspected or estimated cases from credible sources, it's likely this story would have been taken more seriously and not just by Rosie O'Donnell. It may turn out that our dogs and cats were the canaries in the coal mine of an enormous system failure -- one that could have profound impacts on American food manufacturing and safety in the years to come.
I used the "canaries in the coal mine" analogy a week ago. Based on what little we know to date, the death toll must be in the thousands and the rate of renal injury must be far higher. The emotional cost to tens of thousands of Americans is very high, and they will not forget that their government, yet again, failed them. Anyone who doesn't think this couldn't have happened to human consumers is sadly lacking in imagination. Besides a failure to comment on the likely injury rate, another item not covered in this otherwise topical review is the Homeland Security angle.

Once again, we have legions of obsessive bloggers to thank for keeping a story alive that the mainstream media keeps trying to forget.

Oh, and yes, Rosie O'Donnell's producers and support staff are either very dim or very nasty.

Clearwater: crushed by condominums

We're heading back from a semi-spontaneous family spring break trip to Bradenton Beach Florida (Tradewinds Resort - recommended). Great trip, though things will look different a year from now, and even more different five years from now. Only global warming and hurricane risk can keep the condos away.

Clearwater Beach is exhibit A. It's still a beautiful sight heading out from land, but the condo density is amazing. They're mostly dull rather than ugly, but there are a lot of them. I assume the economics of such high density, high cost accommodation are pretty overwhelming, in time they'll swallow the Keys -- unless the storm threats become overwhelming. Of course the ultra-rich will still be able to build estates and hang the cost -- but even in America this is a relatively small group.

I suspect most folks prefer the old Florida of single story wood frame buildings. We found a fair bit of that on Anna Maria island, but it felt like an anachronism.

Besides condos and estates, one other category of dwelling caught my eye. Southwest Florida has the greatest concentration of genteel trailer parks I've ever seen. I assume they're mostly occupied by non-wealthy elders. I wonder how they fit into the evolving land use pattern, and where the economics of the florida mobile home are going to go
... one out every five new single-family housing units purchased in the United States is a mobile home, sited everywhere from the conventional trailer park to custom-designed "estates" aimed at young couples and retirees...

Lycra swim shirts (rash guard) - no sun-fearing geek should be without

After five days of Florida sun, I adore my Men's Long Sleeve Outrigger Rash Guard (lycra) shirt from Lands' End. At $36 it's not cheap, but I'll wear this one until it falls apart. You can buy similar shirts on Amazon for about $23-$35, I suspect they work just as well. It's surprisingly comfortable when dry in the mid-80s, and when wet it gives me a few comfort degrees in the water and cuts air heat completely.

The shirt had no SPF rating, I've seen them quoted as "18", but this SPF stuff is nonsense. If I'd been wearing SPF 50 sunblock I'd have been disabled by the third day of nonstop sun. I ended up as pale as a dead fish in the area covered by the shirt (I did use sunblock beneath it). I paired the shirt with my ultra-fashionable camping hat -- super broad bill, rear fabric drape from hairline to shirt top. Works fine in the surf, and it's great at keeping beautiful starlets at bay (the latter is a joke, the less fashionable beaches we visit show why American endocrinology has such a great future).

It's true that manly men and Republicans have a hard time wearing these shirts, though I think they may be popular among surfers. Geeks and Democrats, however, should consider them proof of a rock solid ego. Recommended.

Update: For next year's model I want a hoodie. A Lycra hood I could slip on and off would work better underwater than my camping cap ...

Sunday, April 01, 2007

Aminopterin, melanine and what else?

We don't know much about what's likely killed hundreds, or thousands, of us pets. Aminopterin could not be isolated from pets,
Melanine is in the Chinese gluten, but it's not supposed to be nephrotoxic. The focus is still on the Chinese gluten (which apparently has lots of fun ingredients), but maybe it's not the gluten after all. A dry cat food has also been recalled.

I hope Homeland Security is paying attention.

In the meantime, my personal conclusion that our family needs to switch to human-regulated foods for our dog stands.

Saturday, March 31, 2007

A clear flaw with closed source: longevity

For longevity, one needs unencumbered (no patent issues) file formats at a minimum, and probably open source code ...
MacInTouch 3/07: Archival Issues (reader contribution)

... One of my fondest treasures is a 32 disk CD-ROM set in a mahogany box, "The Complete National Geographic" magazine. It was a searchable collection of every page of every issue of National Geographic for 109 years. I bought it in system 7.5 days and used it up thru OS 9. But support for this collection met a tragic death

National Geo had farmed out the development to a chain of third party developers...Mindscape, The Leaning Center, and Broderbund (of PrintShop fame) whom for whatever reasons had "difficulties" in sustaining the project regardless of its merits. Each company's edition under the National Geo brand came with a proprietary Reader/Searcher. The various versions were not compatible, not even in consecutive years... The series died about 1999 and support for it didn't last much longer. Some of those companies still exist but avoid all talk of the Complete National Geo debacle. I've tried. Even National Geographic customer service is curiously mute.

It's exasperating to think that all that historical data, all those articles, all those photographs, are sitting on my shelf and cannot be viewed with today's operating system. ( The Reader/Searcher looked to my naive eye like a kissing cousin to Acrobat.)

If only the source code for the Complete National Geographic CD-ROM set were available and could be updated to run natively on OS-X and other contemporary platform...
Let's learn from this. Don't invest in products dependent on closed source solutions. Speaking of FairPlay and every other DRM solution ... What do you think the chances are you'll be able to read a FairPlay DRMd file in 10 years?

Shame on National Geographic, btw. The best explanation is they didn't negotiate their contracts properly ...

Wednesday, March 28, 2007

Apple TV ships, iTunes Store a smoldering crater

Is it coincidence that within a few days of Apple TV reaching consumers the iTunes Store is unusable? It's not quite dead, but it's too slow to buy from.

An ominous hint of what consumer video distribution will mean to the net? Not all bad news for Apple though ...

Kristof on how to turn lunch money into hope

Kristof, who never gives up, has an article on how micro-finance has become personal. He visits his business partners in Afghanistan...
The New York Times
March 27, 2007
You, Too, Can Be a Banker to the Poor
By NICHOLAS D. KRISTOF

...From my laptop in New York, I lent $25 each to the owner of a TV repair shop in Afghanistan, a baker in Afghanistan, and a single mother running a clothing shop in the Dominican Republic. I did this through www.kiva.org, a Web site that provides information about entrepreneurs in poor countries — their photos, loan proposals and credit history — and allows people to make direct loans to them...

... Mr. Abdul Satar said he didn’t know what the Internet was, and he had certainly never been online. But Kiva works with a local lender affiliated with Mercy Corps, and that group finds borrowers and vets them.

The local group, Ariana Financial Services, has only Afghan employees and is run by Storai Sadat, a dynamic young woman who was in her second year of medical school when the Taliban came to power and ended education for women...

...Web sites like Kiva are useful partly because they connect the donor directly to the beneficiary, without going through a bureaucratic and expensive layer of aid groups in between....

...A young American couple, Matthew and Jessica Flannery, founded Kiva after they worked in Africa and realized that a major impediment to economic development was the unavailability of credit at any reasonable cost....
Yes, it's a bit gimmicky, it would presumably be more efficient to give the money to CARE.ORG and let them donate to a microfinance organization. Still, the human angle might help draw in some extra funds. Going forward, the interesting possibility is going the next step, and becoming an active business partner. So make the loan, but also become involved in the enterprise. A bit like being an armchair peace corp person.

I'll give it a try.

Aminopterin update: likely thousands harmed

There's not much coverage now, but Google picks up this update. I will wager, based on the VIN reporting, that the number of animals harmed is in the mid to upper thousands.
ABC News: Group Says Pet Food Deaths Underreported

ALBANY, N.Y. - At least 471 cases of pet kidney failure have been reported in the 10 days since a nationwide recall of dog and cat food and about a fifth of those pets have died, a veterinarians' information service said Tuesday...

... Paul Pion, founder of the Veterinary Information Network, which counts 30,000 veterinarians and veterinary students as members, said Tuesday the number of reported kidney failure cases had already grown higher than the 471, but he said he wouldn't have an updated tally for a few days.

Of the reported cases, he said, 104 animals have died. The network's survey results were earlier reported by the Los Angeles Times.

Pion, a California veterinarian, said only 10 percent to 20 percent of the people who belong to his Web site had responded to a request for information.

... Researchers at the New York food lab, Cornell University and other labs were still working Tuesday to pinpoint which individual ingredients were tainted with the poison, officials said. They also said there could still be undetected hazards in the food...

The labs are having trouble working with the gluten, so it will be some days before they can determine if gluten was really the source, and if the contaminated gluten is entirely Chinese. They are careful to note they cannot exclude other toxins. If we ever understand how the Aminopterin contamination occurred, we will have a better idea of the likelihood of secondary toxins.

A conservative extrapolation of Pion's numbers suggests a renal failure toll to date of about 4,000 pets, with a death rate still running at around 20%. It is likely the majority of the sick had more vulnerable kidneys, probably older cats. A larger number of animals will suffer significant kidney damage without symptoms. Their lives may be shortened. It is likely that significantly more than 4,000 animals will have been harmed.

When we learn how this happened, we will also be able to draw inferences about how often lesser problems may have occurred with pet food manufacturing and distribution over the past few years. It would be surprising if all the problems are being detected; error analysis in other domains always finds that these "sentinel events" represent the "tip of the iceberg".

We will continue to investigate if we can substitute human-regulated food for a significant portion of our dog's diet going forward.

Tuesday, March 27, 2007

Fear in the blogosphere: Microsoft hunts again

I suspect there are a few young techies who do not fear Microsoft. They don't remember when Gates et al pillaged the landscape, sowing salt on the burning embers of resistant villages, conquering with ruthless power and endless treachery. I confess, I have contributed to a false sense of security by pointing out how incompetent Microsoft has been in the past few years.

It is time for me to right my wrong. Microsoft hunts again.

There is only one good tool on any platform for efficient blogging in both Google's Blogger and the open source Community Server (metaweblog API) environment, and it's Microsoft product - Windows Live Writer. I've tried all the rest, they don't compare. [See Update!]

Live Writer only works with IE. It only works with Windows. Yes, there's a Firefox plug-in for Live Writer. It broke long ago, and, oddly enough, it's never been updated.

BlogJet was very good in version 2.x, but version 3.x is a mess. Scribefire is very minimalist and just broke with Firefox 2.03. One the Mac side Mars Edit and Ecto aren't even in the race -- they assume that all writing is done from a single machine. They can't properly handle fetching and editing past posts. They are also updated slowly if at all -- I could go on. They are goners.

Google's integrated "BlogThis!" bookmarklet hasn't been updated to support Blogger 2.0's tags and it's slow, quirky, and underfeatured (Blogger 2.0, btw, is a vast improvement over Blogger 1.0).

So Microsoft has a free tool that is vastly better than anything else, it only works with IE/Windows, and it's smack in the middle of a strategic market. Wow, doesn't this feel familiar?

Is anyone awake out there? Have all the geeks been lulled to sleep by Microsoft's seeming incompetence over the past few years? Where's Google anyway? What happened to all those Firefox development efforts? Is Apple trying to ignore the blogging world? (Anyone remember the transient blogging tool once bundled with new Macs?)

All is not lost. Sharepoint 2007's blogging support is miserable, and Microsoft has invested a lot in Sharepoint. Maybe the Sharepoint team will end up destroying Live Writer. I know I'm rooting for them, because as much as I like Live Writer, I've seen this movie before ...

[1] Update 1/5/09: Thanks to very helpful comments on this post I have learned that Adobe Contribute for OS X and Windows (list $170) will work with Blogger and other standards compliant platforms. It's very expensive and suffers from being an Adobe product (they don't seem to understand how to install OS X software) but I'll try to get a look at it.