Gordon's Laws for software and service use

CrashPlan gets great press and even a Tidbits Take Control recommendation, but when I used it I ran into numerous fundamental flaws. Clearly, I can't rely on reviewers.

From that and similar experiences, here are Gordon's Rules of Engagement for software and services.

Desktop software

1. Is there obnoxious DRM? (Some DRM is understandable, but it shouldn't be obnoxious.)
2. If distributed on CD, can the product be used without the CD running?
3. Look at the installer. Drag and Drop is fine, but if it needs an installer it better be Apple's installer.
4. Inspect the uninstaller. The best apps don't need one - just delete the app. After that look for something built into the app. Then look for something that downloads with the app. If there's no installer stop immediately.
5. If it's software, is there an full feature trial period? Limited feature trials are worthless. I need at least a month, or, better, 10 days of use (which may take me months).
6. Who makes the product? What's their support site like? Can you find downloadable fixes?

Cloud services

1. Is it obvious how to delete your account and all data and services?
2. Do they want your Google credentials? If so, run and bar the door.
3. Do they support Oauth? Do they allow you to have multiple Oauth credentials associated with your account? Extra points for each.
4. Do they require a security question? If so, they're stupid. (Yes, even Google is a bit stupid these days - but they don't REQUIRE it.)
5. If your storing something precious online (ex: backup data), what's the password reset policy? "Industry standard" practices means losing control of your email will cost you ALL your backup data. (for example)
6. Can you get your data out in a useable way? If not, run, run, run.
7. If there are annual renewals, is there an option to request approval prior to renewal?

Desktop or Cloud

1. Is there a high quality manual and/or help resource? It doesn't matter whether you're going to read it or not. Products with good manuals are almost always good products. It's a very reliable quality measure.
2. Is there a blog? Are the developers proud of their work?

Notice there's nothing in here about features, reviews, price, performance, etc. They only matter if a product passes the above screening tests. In fact it's rare for a product to pass all of the relevant tests and then be fail due to bugs or performance. A vendor who can do the above can usually do the product as well.

See also:

• Gordon's Notes: Requiem for a gerserker - Gordon's Laws of Geekery
• Gordon's Notes: Gordon's 4 rules of acquisition

Macroscopic quantum mechanics

In Greg Egan's Teranesia [1] one story I can't currently locate (h/t Mel Anderson, comments), the protagonist is fighting the ultimate infection. It seems impossibly mutable. Turns out it has evolved to exploit quantum effects, and it's finding the perfect mutation by exploring all the many worlds of variation.

That wasn't the only science fiction story of the past decade to imagine that biological organisms, operating at atomic scales, might exploit quantum effects. Alas, science fiction memes don't last long these days. Protein exploitation of quantum effects has become a mainstream research topic. This Nov 2009 Sci Am news article is a good overview of the underlying physics; note especially the resolution to the old debate about how the quantum/classical transition happens ...

How Noise Can Help Quantum Entanglement: Scientific American

... In the modern view that has gained traction in the past decade, you don’t see quantum effects in everyday life not because you are big, per se, but because those effects are camouflaged by their own sheer complexity. They are there if you know how to look, and physicists have been realizing that they show up in the macroscopic world more than they thought...

... This work suggests that, contrary to conventional wisdom, entanglement can persist in large, warm systems—including living organisms. “This opens the door to the possibility that entanglement could play a role in, or be a resource for, biological systems,” says Mohan Sarovar of the University of California, Berkeley, who recently found that entanglement may aid photosynthesis ... In the magnetism-sensitive molecule that birds may use as compasses, Vedral, Elisabeth Rieper, also at Singapore, and their colleagues discovered that electrons manage to remain entangled 10 to 100 times longer than the standard formulas predict...

A quick search on scholar.google.com finds many references on how quantum effects might alter molecular behavior in neurons.

It's a small, small world after all.

[1] See also: Mind expanding books: a list and my comments there on Egan's Incandescence.

--
My Google Reader Shared items (feed)

Do knowledge workers burn more cerebral calories?

Our brains burn a lot of the calories we eat [1]. So are knowledge workers working out?

Alas, probably not ...

Appraising the brain's energy budget (PNAS 2002)

... In the average adult human, the brain represents about 2% of the body weight. Remarkably, despite its relatively small size, the brain accounts for about 20% of the oxygen and, hence, calories consumed by the body (1). This high rate of metabolism is remarkably constant despite widely varying mental and motoric activity...

Later articles suggest that while brains use a lot of calories, and are thus a very expensive evolutionary development, they don't use the calories for thinking. Brain calories are primarily consumed in "intrinsic activites" unrelated to environmental stimulus -- presumably maintenance functions of some sort.

So we knowledge workers don't get any caloric credit for thinking, and since we usually think while sitting (very bad) we're really pro blubber.

[1] We also know human brains are smaller and probably more efficient than they used to be. Of course, so are computers.

Update: I suppose memory formation might be energy intensive, so maybe forming more memories could burn calories?

St. Paul (MSP) AT&T iPhone voice service now intolerably bad

Something has changed over the past year with our iPhone voice services around our home (Macalester-Groveland, St Paul, Minnesota) and, to a lesser extent, along my commute (St. Paul to Roseville).

Voice services have gone from mediocre to intolerably bad. The “bars” are meaningless; even with 3-4 “bars” connection failures and call drops are ubiquitous. We can’t reliable make or receive calls from our home, which is rather a problem since we don’t have long distance landline service.

Curiously data services (3G) are doing well. We’ve tried turning off 3G services at home (EDGE only) to see if voice connections improves, but that doesn’t help enough. I suspect local AT&T carrying capacity is the problem, not wireless signal.

We rely on our iPhones for a lot of things, but we need voice services.

I understand that iPhone-class technology, and the absurdity of flat-rate data service pricing, has put a great deal of strain on AT&T’s networks. I understand that our community is resistant to installing new cell towers. That understanding does not translate into sympathy. Our family is paying, I’m chastened to confess, thousands of dollars for services AT&T is not delivering.

There are things AT&T could do. They could end their insane flat-rate pricing, and institute pay-per-use bundles with similar value but user-aligned incentives. They could provide us with a free AT&T 3G MicroCell (aka femtocell) along with discounts for use. They could give us substantial discounts on their mobile services pending a fix.

They’re not doing any of these things. Instead of providing free MicroCells and discounted services, for example, AT&T charges for their femtocell solution.

I miss the days when it was possible to initiate class action lawsuits for failure to deliver contracted services.

If some other company gets iPhones in June our family will switch. We have only one under-contract iPhone and we can sell it and pay the AT&T penalty.

If no other company gets iPhones, I will find out how good Droid really is.

Update: A computer generated (location customized) AT&T response has some relevant details:

… We see that the coverage around your home is considered to be our best coverage and it includes 3G service.   We also see that there is a planned tower about 2 miles from your home at Osceola and Lexington Parkway S.  This tower is slated to be operational mid August 2010…

…Please contact Customer Care at 611 from a cell phone or at 1-800-331-0500 from landline phone if the problem persists.    Have your wireless device available to allow for proper troubleshooting if problems persist…

So even though our coverage is failing on multiple phones, AT&T considers it to be pretty good. There’s something wrong there.

The 3 miles tower is too far away to help us directly, but it’s close enough to reduce the burden on our proximal towers. On the other hand by August we’ll have the 2010 iPhone and the 3G iPad – so any additional local AT&T capacity will be swamped.

If you do try to phone the 800 number, check the gethuman recommendations:

Press "1" at the system asking for your phone number, and then press "0" at next prompt.

Things are unlikely to improve.

The new history is deep history

When we think about science, most of us think of dramatic breakthroughs. We think Darwin and Wallace, Einstein and Bohr, Copernicus and Curie and we imagine everything changed overnight.

Most science, however, develops in bits and pieces, twisting and turning, waxing and waning, until, after thirty years, things are new. Even the dramatic shifts, like natural selection, took decades to get from radical to mainstream.

If you’re at all curious about things, you notice this in a single lifespan. Consider deep history; the story of humans from 150K to 3K years ago. In the past 30 years discoveries from genomics, climate research, linguistics, plant research, translation, anthropology and archaeology, combined with the revision of old biases, have dramatically changed our understanding of deep history. In each case, of course, computation has been a fundamental driver. That’s how it works – new instruments make new science.

It’s been growing slowly from all directions, but the sum is a very different world from what some of us learned in the 1970s. The human brain is evolving and changing far more dramatically than we imagined, and that evolution has not slowed with modernity. Our concepts of human speciation are being transformed; there were many “species” of human coexisting into deep history – and, like dogs and wolves, they probably crossed often.

Pre-agricultural humans were far more populous and widespread than we once imagined; the large populations of pre-invasion (early agricultural and hunter-gatherer) North America probably reflect worldwide pre-agricultural patterns.

Even after the development of agriculture and writing we see thousand year intervals of relative stasis in China, Egypt and Mesopotamia. How could this be when our fundamental technologies change in decades. Are the minds of modern Egyptians radically different from the minds of only 6,000 years ago? Why? Why do we see this graph at this time in human history?

What did humans do in Georgian caves for 30,000 years? Thirty thousand years of waving and sewing and nothing changes?! They could not have had the same brains we have. They seem more … Neandertal.

Fascinating times, and there’s much more here than I can address in one post. That’s why I’m adding a new tag (label) for this blog -- “deep history” in anticipation of much more to come.

For now see also:

• Dienekes' Anthropology Blog- Paleolithic flax fibres from Georgia (Kvavadze et al. 2009)
• Multiregional evolution lives! - john hawks weblog
• Why Humanity Loves, and Needs, Cities - Economix Blog - NYTimes.
• In Our Time, Babylon
• In Our Time, Akhenaten (He reminds me of Kim Jong-il)
• In Our Time, The Library at Nineveh
  

Update: What does the Antikythera mechanism teach us about deep history? It cannot be the only anomaly of its kind in all time.

Update 8/21/2010: Brad DeLong says I don't pay enough attention to the fragility of culture and technology without writing and location persistence.

How bubbles destroy software - the story of Broderbund

My kids need some typing software. I remember a "Broderbund" "Mavis Beacon" product from eons past, so I started looking into the current state of the product. It took me a while to sort things out; and along the way I was reminded of how much was lost in the tech bubble of the 1990s.

It turns out there's now a Mac-only "Mavis Beacon" product sold by "Software MacKiev" (Ukranian Mac contract software development) and another product (XP/Mac) sold by "Encore Software". The Encore Mac product is buggy and unsupported, the MacKiev version sounds a bit more promising.

it seems the current products are both descendants of a common ancestor that passed through a number of corporate labels including Software Toolworks / Mindscape, Broderbund, and The Learning Company .

What a mess. How did that happen? For that matter, what happened to all of the pretty good educational software produced by all of those companies? How did it all die, without true replacements (the closest things today are Flash apps on commercial sites that do child marketing)?

Part of the answer is that these companies had a good skillset for computer gaming, and that was a much bigger industry. Another part of the answer comes from the the wikipedia article on Broderbund (1980s history here) [1], (emphases mine, remember those who got cash were the winners) ...

... Softkey ...purchased The Learning Company for $606 million in cash and then adopted its name...Brøderbund was purchased by The Learning Company in 1998 for about US$420 million in stock...

In a move to rationalize costs, The Learning Company promptly terminated 500 employees at Brøderbund the same year,[16] representing 42% of the company's workforce.

Then in 1999 the combined company was bought by Mattel for $3.6 billion ... Jill Barad, the [Mattel] CEO, ended up being forced out in a climate of investor outrage.

Mattel then gave away The Learning Company in September 2000 to Gores Technology Group, a private acquisitions firm, for a share of whatever Gores could obtain by selling the company. In 2001, Gores sold The Learning Company's entertainment holdings to Ubisoft, and most of the other holdings, including the Brøderbund name, to Irish company Riverdeep.[19] Currently, all of Brøderbund's games, such as the Myst series, are published by Ubisoft...

This kind of churn is death to software. Software needs continuity to survive. The cycle of acquisition and 'rationalization" creates zombie software that staggers on, brainless, for years ... then dies.

The tech bubble made a few people rich, and it destroyed a lot of good products. Not to mention costing Mattel's shareholders quite a few pennies.

After the tech bubble burst came 9/11, then the great asset bubble and, not least, the Bush administration. One, two, three, four. No wonder America is reeling.

[1] At one time Mavis Beacon was sold under the Broderbund name, but by that time Broderbund might have been owned by Riverdeep. I include this story as an example of all the things the tech bubble killed.

Update 4/30/10: I bought the MacKiev product. It came with a solid, richly photographed manual with the name "Broderbund" on the front cover. This team is proud of their work. The manual included some interesting background on how Software MacKiev ended up doing their own OS X version:

Mavis Beacon Teaches Typing was created more than twenty years ago, and was first published in 1987. Software MacKiev’s involvement goes back to 1998 when our company developed version 9 for the Macintosh — both the US and UK editions. Then, a decade later, we had the opportunity to get involved with Mavis Beacon again — this time as the developer and publisher of a new generation of Mavis Beacon software for Mac OS X. We are so pleased and proud to be bringing the kind of quality you’ve come to expect from the creative labs of Software MacKiev to this new edition.

Update 6/8/2010: I've received a few more details. The "Encore" versions of Mavis Beacon for OS X are arguably fraudulent. They're not designed for the current OS. Looks like "Encore" bought up some discarded software assets ...

... Software MacKiev develops and publishes only the Mavis Beacon Teaches Typing 2008 and 2009 Deluxe, International, and School Editions for Mac OS X. The previous versions of Mavis Beacon were made for really old Macs with OS 9 by a company called Broderbund. A company called Encore has since taken over the broderbund.com Web site and continue to distribute the outdated software, which — as you point out — doesn’t work as it should...

Exercise, weight loss and dementia

It's not that exercise doesn't work as a weight loss method, it's just inefficient ....

Weighing the Evidence on Exercise - NYTimes.com:

... When researchers affiliated with the Pennington center had volunteers reduce their energy balance for a study last year by either cutting their calorie intakes by 25 percent or increasing their daily exercise by 12.5 percent and cutting their calories by 12.5 percent, everyone involved lost weight. They all lost about the same amount of weight too — about a pound a week. But in the exercising group, the dose of exercise required was nearly an hour a day of moderate-intensity activity, what the federal government currently recommends for weight loss but “a lot more than what many people would be able or willing to do,” Ravussin says."

An hour a day would be wonderful -- if my kids were grown.

The NYT article has the complex details. The effect of exercise on weight varies by age and gender, and between individuals as well. In general, however, it's not a good way to lose weight. Diet is more efficient.

On the other hand exercise seems to be essential to keeping weight stable after a weight loss diet. How and why? Nobody knows for sure.

Sitting turns out to be really, really, bad. We've had hints of that over the past years, but now it's getting pinned down. We don't know why, but sitting promotes obesity.

Incidentally, it's all harder for women. But you knew that.

Elsewhere in the NYT, Olivia Judson deepens the cheer with claims that obesity causes brain damage. She's a bit below her par though; she obscures correlation with causation in the interests of more hits(she well knows the difference, so two demerits to her). It is likely that obesity is associated with early dementia, but it's also associated with lower socioeconomic status, lower IQ, and the anger of the gods.

For a contrary opinion, see Gina Kolata (also NYT) in 2007.

On the other hand, there's a weird association between exercise and brain function, even though I didn't believe it years ago. Exercise seems to help the health of neurons associated with cognition and memory in various animals -- for no particularly good reason. Since exercise is associated with lower obesity (in both directions) this further murkens the muddies.

Exercise also seems to help sleep, and I do suspect that sleep will turn out to be very important for brain health. Since obesity does impact sleep quality there may be an effect of obesity on the brain both by diminishing interest/ability to exercise and by worsening sleep quality.

Lastly, the idea that brain activity (bridge, crosswords, etc) slows dementia seems to be, at long last, good and dead. It doesn't work. Forget the bridge, forget the crosswords, go for a walk.

To sum it all up, my best guess at how this will all turn out:

• Sleep is more important for brain health than we've imagined.
• Exercise is more important for brain health than I thought 4-7 years ago. (I like to exercise, so it wasn't a prejudice against activity. It's just weird science.)
• Exercise helps both sleep and brain health - so it's a double good. It doesn't lead to weight loss, but it's essential to maintain a stable weight.
• We all need to diet all the time, so we need cultural and industry changes to make that very hard activity easier.
• Obesity is almost inevitable in a food rich world, especially when we eliminate smoking and increase sitting (at computers). We need a miracle drug, we need cultural changes, we need mobile devices, we need gas to hit $10 a gallon.
• Sitting is oddly bad for us. We should all be standing and walking.
• Try not to get a concussion (but almost all enjoyable exercise increases head injury risk :-). Don't let your kids play football (which will eventually go the way of boxing).

See also:

• BBC NEWS | Health | Creatine 'boosts brain power' 2003: Right. This went nowhere. A good reminder of how worthless most press releases are. We're still hearing about the huge onslaught of cognitive enhancer abuse, but it's mostly media imagination.
• Obesity increases dementia risk: Winner of stupid medical article of the month 2005: Early reports on the obesity/dementia causal relationship. I'm very skeptical then. I still am on the obesity side of things.
• Exercise and Alzheimer's: How to fix the demented media 2006: I really didn't believe this particular correlation, but since then it's shown up in more persuasive animal studies.
• Alzheimer's: there's not much you can do about it 2006: Head injury avoidance remains important -- but also lifestyle constraining. We all die one way or another!
• Exercise and cognition - I'm still skeptical 2007: I'm starting to wonder if I was wrong to dismiss a causal relationship between exercise and brain health
• Skinnies burn fast, plumpies less demented 2007: In this study the elder skinnies became demented faster than the pleasingly plump.
• Delaying the inevitable onset of Alzheimer's: exercise is the new best hope 2008: When I admin that it does look like exercise is oddly good for brain health.
• Wasting money: Ginkgo biloba joins the pile 2008 - Anyone remember when Gingko biloba was going to help dementia? Fail.
• Dementia is normal - and what that means: 2009 - We're beginning to realize dementia really is inevitable -- though there's still hope to push it back. (Death and homeliness are also inevitable)
• Avandia (rosiglitazone) for Rat Alzheimer's Prophylaxis?! 2006 - It didn't do too well in human trials. It also promotes obesity, which is kind of interesting.
• Dietary supplements stop dementia in beagles? AAAS proceedings 2004 - Nothing came of this!

Hysteresis

I lost the word "hysteresis" years ago, and I've been looking for it ever since.

It can be used to describe the common situation where a sustaining force is removed, but a dependent state persists -- for a time ...

Hysteresis - Definition and More from the Free Merriam-Webster Dictionary

a retardation of an effect when the forces acting upon a body are changed (as if from viscosity or internal friction); especially : a lagging in the values of resulting magnetization in a magnetic material (as iron) due to a changing magnetizing force

I knew the concept, but Google refused to resolve my descriptions into the word. I had no choice but to wait until someone used it. That took years! I just saw it in a Sci Am article on how a newly frail antarctic ice mass responds to climate change.

I keep a short list of the handful of odd words I tend to lose (like gratuitous and vicarious), now hysteresis has a place of honor.

Electronic health record use and physician multitasking performance

"LLamas" has a pithy, if unfortunately unreferenced, summary of how most brains fail when they're trying to multitask ...

Llamas and my stegosaurus: Living with a limited brain

Some interesting research has come out recently about the processing capacity of brains. For example, that the medial prefrontal cortex can only handle two tasks at once, or that working memory can only handle about 7 items at a time (but what's an item?), or that when people are actively trying to remember something complicated, their impulse control is reduced. In fact, there has been a lot of research showing that exerting the will to make a difficult decision uses a fuel resource (sugar from the blood) that many of these other tasks also need.

What happens when these resources are used up? When we have been thinking too hard, or have been under heavy stress, or haven't had enough to eat or sleep, or are trying to remember too many things, or are trying to drive, or need a fix,we fall back on a simpler part of the brain. We lose the ability to think rationally, to choose future benefit over immediate reward; the ability to choose at all is reduced. We become irritable, forgetful, angry, quick to argue....

I've been disappointed that there have been few studies of how physician cognition adjusts to using automation tools (electronic health records, etc) during patient care. These tools all seem to have a substantially higher cognitive burden than phone use, but the impact of phone conversations on driving performance has been studied to death. Do physicians become more irritation and distracted when they try to simultaneously talk with patients, think about the answers, and use current clinical software?

Cloud printing - 1994

Google wants to do cloud printing for the ChromeBook. The iPad can't print either.

That feels familiar. Where did I hear about PostScript style cloud services before?

Oh, Yeah. TeleScript - 1994. It was a Xerox Park/General Magic/PostScript inspired kind of cloud services thingie.

I suspect IBM mainframe printers did something comparable once upon an eon. Moving bits of the OS between client and server.

Myself, I'm looking forward to embedded OS scanners that will let me pick up the scans off the network. That's the way the big office machines work, and I'd love to separate my home scanner from my machines.

If computer technology had developed like health care technology

If computer technology had developed like health care technologies home computers would cost millions. Everyone would have one, but the nation's computing insurance company would be facing disaster.

See also: Staggering towards health care access ...

xkcd spookiness

xkcd is reading my mind (click to enlarge) ...

Spooky. I am one with xkmind. Except that not only can I wait, I would really prefer that the future spare my children too.

The xkcd story is pure 21st century geek culture. How geek? On Google Reader xkcd gets over 100 "likes" within hours of release. That's probably the fastest and highest "like" rating in the entire geek universe.

How geek? Here's the official xkcd reuse permission policy ...

xkcd - A webcomic

... If it's a not-for-profit publication, you need no permission -- just print them with attribution to xkcd.com. If it's a for-profit operation, I will probably give you permission if you email me to let me know. You can post xkcd in your blog (whether ad-supported or not) with no need to get my permission...

Munroe makes money by selling xkcd paraphernalia and a book of favorites ...

... In September 2009 Munroe released a book, entitled xkcd: volume 0, containing selected xkcd comics.[80] The book was published by breadpig, under a Creative Commons license, with all of the publisher's profits donated to Room to Read to promote literacy and education in the developing world. Six months after release, the book has sold over 25,000 copies. The book tour in New York City and Silicon Valley was a fundraiser for Room to Read that raised $32,000 to build a school in Laos....

Update: I corrected the original post. Munroe's publisher has donated the book profits, but Munroe can keep his well deserved share.

My Google Reader Shared items (feed)

Google Chinese to English - a rare good result

Google translate is built into Google Reader. That allows me to follow someone who shares both English and Chinese material.

Google's translation of the Chinese content is usually unreadable. Each sentence seems more or less correct, but the sum is incoherent.

Today, however, I came across a post I could read. The sentences were still a bit stilted, but the paragraphs were coherent. Overall it read like a bilingual native Chinese speaker writing in quite good English.

The trick is that the interview was conducted in English, then human translated to Chinese, then Google machine (statistically) translated back to English. The Chinese translation must have preserved quite a bit of the original sentence structure; enough that the reverse translation worked quite well.

Fascinating!

--
My Google Reader Shared items (feed)

Weingarten's accidental infant death story - now with a Pulitzer

The Weingarten story on infants who die when accidentally left in the back of a car was published in March of 2009.

I didn't read it then. Unlike Ray Morrogh, an astoundingly arrogant and ignorant prosecutor, I knew this could happen to anyone. I knew it could happen to me.

I've studied human cognition and human error. I knew that the only way to prevent these disasters is to reengineer car seats and car systems. [1]

Today Weingarten won the Pulitzer, and, more or less by accident, I read Fatal Distraction: Forgetting a Child in the Backseat of a Car Is a Horrifying Mistake. Is It a Crime?.

I wept at the end. Surprised me, since there was nothing in there that was new. When we moved car seats from the front seat to the back seat we saved many lives, but we made these errors inevitable.

It is very well written.

[1] If I still had infant passengers, I would clip a lead from the car seat to my belt every time I got in the driver's seat. Then I'd have to leave my pants in the car to forget the infant. I only heard of that fix after my kids were mobile.
--
My Google Reader Shared items (feed)

What iPhone multitasking can do

There are quite a few iPhone apps that are too slow now. Appigo’s ToDo.app needs to sync with tooldedo. That takes too long, I hate it.

Evernote needs to sync and/or upload. Way too slow.

There are other apps that are incomplete, because making them complete would make them unusable. I’d like my Voice Memos to be transcribed, but I want zero delay on launch and dictate.

In all of these cases the time consuming task doesn’t need to take place while I have the UI in front of me. They can take place after I leave the app.

Multitasking solves all of the above. The sync/communicate tasks can occur in the background while I move to do something else.

That’s why I want iPhone multitasking. We’ll see in the fall if Apple’s version of limited multitasking is good enough. I fear it will be too limited for another year.

Update 4/14/10: Good news. The functions I want fall under the "task completion" domain. That should be available to all apps.