Comments by D. Sherman

Y2K was a very legitimate problem, mostly for old accounting programs that still had decades-old COBOL and JCL at their core. Fixing the bugs brought a lot of old programmers out of retirement, and to their credit, they succeeded at patching the old code.

I knew the hype had reached the point of no return, however, when I heard a "computer expert" being interviewed on the radio who pointed out the alarming fact that nowadays even french-fry cookers and elevators have clocks in them, and that most of them will fail on the transition from 99 to 00. Yes, he was correct that french-fry cookers and elevators have clocks in them, but what was wrong was the implication that anything bad would happen if they got the year wrong. There's nothing about the frying of potatoes or the operation of an elevator that require knowledge of the correct calendar year.

That's why some of this story reads a lot like the Y2K hype -- someone counted how many medical devices have clocks in them, and how many of those clocks are wrong, but apparently nobody eliminated the devices for which the clock function wasn't important to patient safety, nor did they look to see what level of error was harmful for each device. There must be very few devices for which an error of a minute or less would be significant, and yet most errors were less than a minute.

I don't doubt that you are a good engineer at what you do, but if you've been doing it "a very long time" you have probably been doing things that don't involve TCP/IP. Merely connecting a device to the Internet does not open it to malware attacks, and I highly doubt that fear of such attacks is the reason that medical devices aren't networked.
To start with, malware can only be installed if the device includes program memory that can be altered in the field. Devices that are strictly ROM-based cannot possibly be attacked in any way since the memory cannot be written to. More modern and complex devices sometimes permit updating their firmware in the field, but if the design is the least bit competent, doing so will require taking several steps that can only be done in-person at the machine, or at least via some sort of secure network.
Medical equipment doesn't include the most common method of spreading malware, email programs, and it would be very rare if it included a web browser at all. If it did, it would usually be configured only to display very specific html generated internally. Networked devices generally do include web servers, but without a way to alter what is being served, there is no way to install anything maliciously. Obviously, an incompetent engineer with too much time on his hands and too big of a budget could design a blood pressure cuff that included anonymous telnet and ftp, plenty of excess file storage space, an email client, and probably even a game of "rogue", but an incompetent architect could just as well specify that the building's doors not have locks on them.
Competently designed, even without any special expertise, there is no need for any networked medical device to be vulnerable to malware. The only reason PCs are vulnerable is that, by definition, a PC must allow a person to install and run new programs on it.
Rather than going on about the technical details, let me point out that this problem has been recognized, dealt with, and largely solved in industrial control systems. If the idea of a blood pressure monitor getting infected with malware is bad, imagine the entire control system for an oil refinery or a large power plant. In the early days, engineers were naive, and some systems were open and vulnerable, but the basic protections built into all industrial systems these days seem to be more than adequate for keeping out malware. Blowing up an oil refinery is a much more attractive terrorist target than turning off granny's insulin pump, and yet no one has done it.
Networking is the future, and it will ultimately include all but the most trivial medical devices. The benefits are simply too great to ignore; real-time data collection and synchronization, automatic self-test, calibration, and error reporting, and of course the ability to observe and work remotely.

This reminds me of the Y2K scare. During the run-up to Y2K we were repeatedly reminded of all the electronic things that have clocks in them, with the implication that all of them would break if they didn't handle the Y2K rollover properly. As it happened, the important things were fixed in time and the ones that weren't fixed were irrelevant.

In analyzing the implications of mis-set clocks in medical devices, the sheer number of clocks that are set wrong doesn't much matter. What matters is what they are used for and how big an error is significant. I'm an engineer, not a doctor, but I'm an analog engineer, and in the analog world nothing is exact. The question we ask is "how accurate do you want it to be?" "Perfect" is never an option.

A lot of devices appear to have clocks in them simply because it was free to implement in software, especially if the device really did need a timer of some sort. Household examples include all manner of kitchen appliances, laundry machines, etc. I suspect a lot of medical equipment is in this category as well, or if it does "need" a clock, it merely needs it for time-stamping its data record. An ultrasound machine with a mis-set clock is still going to make a perfectly good image of the patient's innards. It just won't print the correct time on it.

Even in cases where absolute time does matter, errors in the range of minutes are probably acceptable in most cases. If it takes several minutes to make an image or run some other test, it doesn't matter if the time-stamp is inaccurate by as much as one minute. On the other hand, there are clearly some cases, including real-time life support, where absolute accuracy in the range of seconds might be vital.

From an engineering point of view, I would separate time issues into three groups -- broken hardware, imprecise hardware, and operator error. The cheapest real-time clock crystals should hold time within a minute per year. That's probably good enough for all but a few very rare applications. If a clock is farther off than that, it was either set wrong or the clock is broken. In some designs it may be possible to include tests for extreme clock breakage within the internal self-test routine. Most plausible electronic failures of real-time clock circuits cause them to fail completely rather than to simply drift out of tolerance. There isn't much cure for people setting the clocks wrong, however, and I suspect that might be the main cause of the hospital synchronization problems mentioned.

In the long run, the cure is to have everything networked and have the firmware synchronize its clock with an internet time server. It could then easily report an error if the internal clock has drifted too far away from that. We can't rely entirely on internet time servers because we need medical equipment to continue to operate even if the network is down, but regular checking is free if the device has a network connection.

I, for one, would like to have read considerably more about exactly how the investment banks "tried to prop up Facebook’s share price." How exposed are they currently? How leveraged is their exposure? Who are counterparties to this risk? Or, jumping to the bottom line, who will lose how much if the price does not rebound soon?

China, the last I heard, was on track to build 50 new coal-fired electric generating plants per year for the next several years. Compared to that, the amount of jet fuel it burns flying into European airports is nothing. All the CO2 ends up in the same atmosphere no matter what the source.

Putting aside for now the environmental issues, it would be nice to read somewhere exactly how "carbon credits" work -- the "transparency" issue Albertican mentioned. In particular, how are the free ones allocated (by gross sales volume, by flight numbers, by annual fuel usage, by fleet efficiency, ?) and who pockets the money for the credits that are sold. Any time the government creates a "product" that never before existed and then sells it (a building permit, or a driver's license, as familiar examples), it's a very neat way to make money, and we'd be wise to pay attention to where that money goes.

Yes and no. You're right that the chemical efficiency of the conversion process isn't great. However in a modern chemical plant, there is no such thing as "waste heat", and much of the heat that used to go up the stack in a 1930s Fischer-Tropsche plant could now be reclaimed, for power generation, process heat, space heating, and even fish farming.

More to the point, what will be needed as conventional oil runs out, is transport fuel. Airplanes will not be running on electricity, or even CNG any time soon, and over-the-road trucks probably won't either. Once you start talking about grid-recharged electric vehicles, you're talking another 50% hit in efficiency relative to the generating plant. What Fischer-Tropsche does is give us a much longer supply of combustible liquids for transport fuel, primarily diesel / turbine fuel, after conventional petroleum becomes too expensive.

Perhaps you could tell us what you've seen. I've never been out of sight of land myself. I've had people tell me there's nothing particularly visible in the "garbage patch". More than likely, it depends on what part you happen to be passing through. It's hard for us landlubbers to comprehend how huge the oceans are. Even milligrams per square meter (if such is the number) quickly adds up to a huge amount.

In my comment, I clearly said that I would have appreciated more information on these two aspects. That's the "research" I would have appreciated the author doing.

From what I've read elsewhere, although there is a huge amount of garbage in the ocean, the ocean itself is also huge, and the "garbage" patch tends to not even be visible when passing through it. We land-lubbers are picturing a sea nearly covered with trash, but that is apparently not the case. The closest I've come is seeing pulverized plastic in seaweed line on North Pacific beaches, but even there there's no visible amount of plastic in the water offshore.

While this might mean it's not as big an environmental problem as it would appear at first glance, it also means that proposals to somehow scoop it up with booms and ships are completely fanciful. It would be interesting to have been told in absolute terms how much there is, in milligrams per square meter, perhaps, and to be told what eventually happens to it all. If natural processes break it down into "micro-trash", presumably the same processes continue and eventually break it into molecules and atoms. Those, of course, might present their own hazards to sea life, but it would be nice to at least know what happens.

The new shale gas technologies and discoveries may do more than any government program to help end the recession. Like all fossil fuels, God isn't making any more natural gas, but this is a much-needed reprieve from the decline in conventional oil.

What I don't understand is why we keep talking about complicated, expensive, and inefficient CNG-fueled motor vehicles, when we could use our abundant methane and coal to make liquid fuels via variations on the Fischer-Tropsch process. Nothing can beat a combustible liquid in a non-pressurized tank as a fuel for motor vehicles, including airplanes. It's pretty easy to make a reasonable diesel fuel using Fischer-Tropsch. Gasoline is a bit trickier, which is why during WWII, the Germans focused on turbine-powered aircraft late in the war. The world is moving from gasoline to diesel anyway, so this is no problem.

So, what's holding up gas liquefication (or coal gasification if you prefer)? EPA rules? Uncertainty about future fuel and raw materials prices? It seems like someone should at least be dusting off those 70-year-old plans and building a gas+coal pilot plant.

A whole lot of Americans ask "Is a Realtor(tm) worth 6%?", and answer "No." Craigslist provides the universal free medium for real estate ads. Ordinary buy/sell agreements can be written and reviewed by a competent lawyer for $500 or so. The legalities of closing -- title insurance, clearing liens, recording mortgages, contracts, and deeds of trust, etc -- can be handled by title companies. There's also nothing wrong with advertising one's property on Craigslist and offering a 3% commission to any agent who brings you a buyer. I've never understood why sellers often state "No agents." in their ads.

There are a lot of real estate agents who, for their 3%, don't do much besides put the property in the Multiple Listing Service. On the other hand, there are some really good ones who get out there and hustle and get the deal done in the face of all kinds of obstacles. Personally, I like a simple deal -- cash for the deed, no contingencies, no inspections, no warranties, and a quick closing -- and am happy to negotiate that myself for the right price, but for complicated deals, a good Realtor(tm) really earns their 3% (for the closing agent -- the listing agent may not of earned their 3%) When you have private parties that are difficult to deal with, buyers with marginal credit, property with obvious "warts", cloudy titles, and so on, an experienced and ambitious closing agent can untangle the morass and get the deal done.

So, the question shouldn't be "Is a Realtor(tm) worth 6%?" but rather, "When is it worthwhile to use a Realtor(tm)?" I find that, despite seller's assumptions to the contrary, a real estate agent usually works best for the buyer. The agent makes their money on volume, not on maximizing the price in one sale. They want to turn property and turn it fast, so if the buyer goes to the agent with a low-ball cash offer, that agent will tend to use all of their skills to persuade "their" seller to accept the offer.

What doesn't make sense to me is why commissions are almost always a flat percentage. I've known agents in small towns sweating blood to earn 6% on houses that were selling for $20,000-$30,000, while agents in more prosperous areas worked no harder to earn 6% on million-dollar sales. If the industry wants to continue justifying its existence, or at least justifying its pay scale, it's going to need to shift to a pricing structure that either reflects the difficulty of doing the deal or at least asks for a smaller bite on the bigger sales.

Please pardon my brevity. I didn't think it necessary to spell out the steps connecting Mr. Krugman's repeated advocacy of monetary expansion and the inevitable resulting decline in the value of the dollar. Perhaps you do not read Krugman, and therefore missed the details of what he regularly pushes for. In his recent piece, as well as almost everything else he's written in the last several years, he claims that the risk of inflation is very low and that most the Fed should increase the money supply. He argues for this as an unmitigated good on its own, apart from what the money might be spent on.

If Mr. Krugman advocated increased infrastructure spending, because the resulting efficiencies in energy, transportation, and communications would improve the economy, and proposed to fund it either by raising taxes or reducing spending elsewhere, we'd have an entirely different argument. I think a good case could be made, for example, that spending $100 billion on nationwide high-speed internet would be better for the economy than spending it on war.

Mr. Krugman argues nothing of the sort, though. His clear point is that monetary expansion is what's necessary, not any particular application of the money thus created. Monetary expansion means dilution of the money supply, which means decreased value of the money already in existence, which means inflation, or as they used to say in the days of Bretton Woods, devaluation of the dollar.

Now you may also choose to argue that the value of the US dollar should fall, that Americans should have to work more to buy less, and that a devaluation would be good for exports, or even for the geopolitical balance of power. Those are all side issues at this point. All I was trying to say here is that Mr. Krugman's latest op-ed is yet another complaint that Mr. Bernanke is not inflating the money supply fast enough.

Mr. Krugman has become a one-note song lately. I don't understand why he even gets so much attention. Everything he writes these days is harping on the same point -- what we used to call "devaluing the dollar", although no one calls it that today. It's getting boring.

Whenever I read an article like this, I remember back to an article about North Korea in National Geographic some years ago that gave a fawning portrayal of the success of communism in that country. The people were all happy, the crops were all bountiful, and the country would soon be self-sufficient in every way. The only problem that the author's guide admitted was that the electric toasters being made locally were not yet quite as good as those made in Europe, though they soon would be. All of this was reported unquestioningly by a National Geographic writer who was clearly delighted to see a shining example of the success of communism, in comparison to the economic inequality and starvation that had been the rule before the revolution.

I do not know if North Korea's electric toasters ever reached the quality of those made in Europe, but that is probably not a big concern to people who can't even remember ever having had a slice of bread to toast in one. Beyond the glaring historical irony of that magazine article, I have to wonder to what extent starry-eyed leftists in the West have, wittingly or unwittingly, helped prop up brutal communist dictatorships around the world. If North Korea is the last true communist dictatorship remaining, it's also one of the most brutal of them all.

At this point, the regime is probably being sustained largely by the desire of China and South Korea to not have to deal with a flood of refugees and the cost of rebuilding a collapsed neighbor, but it would be interesting to know how much of Western policy towards North Korea over the past 50 years has been informed by the desire of the Western leftist intelligentsia to believe that communism must be succeeding there.

I would argue that it's useful to have a national postal service that will deliver a first class letter anywhere for less than a dollar and will deliver mail to any house or business in the country, no matter how remote. It's one of those things that ties a country together, and is probably something worth subsidizing to some extent. UPS/FedEx only provides "universal" delivery by utilizing the USPS for the final leg in remote areas, and if we had to rely on UPS/FedEx for first class mail, they'd likely charge at least $5 to deliver a letter.

There are three things that congress could address immediately, however, to improve the USPS' finances.

The 4th-class "book rate", now "media mail", should be eliminated. This 19th-century anachronism was designed to improve literacy and education on isolated farms, but is pointless today. Mailing books and computer media should cost the same as mailing anything else.

Restrictions on what can be mailed should be repealed. The USPS should be able to carry anything that UPS and FedEx can carry. The quaint prohibition on mailing alcohol is particularly silly since these days there's a substantial mail-order business in wine, and people would often like to mail a nice bottle of wine or liquor as a gift. Instead, all of that business gets driven to UPS/FedEx, along with all the industrial business in paints, chemicals, and the like. Shipping such "hazardous materials" via the private carriers requires a paying a substantial haz-mat fee, which is largely gravy to the carries and could accrue to the USPS' bottom line if they were allowed to carry such products.

Lastly, the USPS' legacy pension obligations need to be addressed. This is a big can of worms, but if congress could deal with it with the railroads and later with the auto industry, they can deal with it with the USPS.

I take it you're not an American. Once upon a time, the Republicans were the party of the rich and the Democrats were the party of the working man. Somewhere along the way, as the Democrats took their long-standing power for granted, they started losing the working people with some of their environmental and social policies that hurt factory and resource workers. Meanwhile, the Republicans hit on the key to getting actual votes, rather than just dollars, which was to become the party of Jesus.
This is a huge generalization, but in general poor people vote Republican because they believe Republicans will enforce what they see as traditional Christian family values. Something like 40% of Americans believe Jesus will return to Earth soon. Don't discount the political impact of that kind of faith, once a party has managed to associate itself with it.

Romney is in no way "on the tea party ticket". He's on the east coast liberal RNC moderate ticket. As for race, what of it? Will Obama's slogan be "Like Romney but only half as white?" You might as well bring up religious feelings regarding Romney's Mormonism.

The reality is that Obama has gotten more conservative (at least in the sense of defense and big business) and has enjoyed his "unitary presidency" and executive orders every bit as much as Bush Junior did. Perhaps he's just the puppet of the Clintons at this point, but the Clintons are nothing if not competent at formulating policy that is largely acceptable. The lack of Republican enthusiasm for Romney stems largely from him being too liberal, especially on socialized medicine. Both men are far too moderate for the extreme wings of their parties. I can't see a huge difference in meaningful government policies no matter who wins. That's why it will be disgusting to watch them accuse each other of radical extremism throughout the campaign.

Once upon a time, candidates campaigned to convince all the voters that they were the best. More recently, they've campaigned to sway an increasingly narrow sliver of moderate voters, since most voters have already made up their minds, if not which candidate they prefer, then which candidate is the very embodiment of evil who must be stopped no matter what. Nowadays, however, with so few voters even in a condition to be persuaded, the campaign is more about turnout. Both parties have plenty of voters to win the election, if they can get them to turn out at a slightly higher rate than the enemy party's voters.

So, what increases turnout? Fear and anger. Hence, we'll be faced with a campaign in which Romney and Obama try to convince us that if we don't turn out and vote (correctly), the enemy candidate will destroy life as we know it. It should be lovely.

It's not just political parties that have their favored words for describing their enemy. Nations pushing a particular philosophy are also at fault. Most recently, of course, the term "terrorist" was more or less given its entire modern meaning as well as a broad new legal definition by the US in promoting a "war on terror" which if examined closely seems to be merely a war on anyone who opposes self-defined "US interests". Domestically, petty criminals are being charged with "terrorism" merely because they terrorized their victims.

It took maybe one year of "monkey see, monkey do" before every tin-pot dictator in the world was describing his own political opponents as "terrorists" and re-branding his secret police, domestic security, and other sundry goon squads as "anti-terrorist" forces. At this point, the word has become meaningless, covering anyone whom any government doesn't like.

"Insurgent" is still a bit more narrow, since it seems to still require that the person or group actually be taking up arms, but in practice it merely means "anyone fighting against our interests", whomever "our" might be. Insurgents whom "we" view favorably might be described merely as "protesters" or "the opposition".

Prior to the war on terror, it was communist governments that excelled at creating their own definition of an otherwise-vague term. Opposition groups, internally, were "counterrevolutionary" or "reactionary", terms that were almost never used in non-communist countries, while troublesome individuals were frequently described by the USSR as "hooligans", a term that never seemed to sound as serious in the West as the Soviets intended.

Within the US currently, the conservatives have so poisoned the words "liberal" and "socialist" that no one dares use them positively or constructively any more, and they're working hard at putting "progressive" in the same category. It took the liberals a little while to catch up, probably because all those bearded college professors wasted too much time trying ineffectually to teach us that the terms were being misused, but they seem to have found their footing now and are throwing misapplied and redefined terminology back at the conservatives as fast as the conservatives throw it at them.

It is as if if all the speechwriters and opinionators have become Humpty Dumpty; "When I use a word, it means exactly what I mean it to mean, no more and no less." More particularly, when they use a word whose true meaning is obscure to the audience, what it means is that their opponent is a dangerously bad man.

"the country needs a “strong and wise government” that can persuade the onsen owners and local communities that the industry would not spoil their spas."
So, the claim is that it's just a matter of marketing to show ignorant people thee truth? One could just as well say, with more megawatts sooner, that the country needs a strong and wise government that can persuade the local communities that the nuclear industry would not spoil their land and their health. A very good case could be made that some modest redesigning and retrofitting could prevent any future Fukushima-like meltdown, that nuclear power is still the best for Japan, all things considered, and that most of the Fukushima "disaster" consists of the government permanently closing off large inhabited areas that would really be harmless to re-enter.
Obviously a lot of people would be skeptical of such a pro-nuclear sales pitch by the government, but they would be no more foolish to be skeptical of that than they would be to be skeptical that the country could switch its current nuclear electric supply to geothermal without damaging the environment or the hot springs.
Both Iceland and Japan have a lot of hot springs, but that's where the comparison stops. Iceland's geothermal resources are higher quality (i.e. higher temperature) and its population density is much lower. That means it needs less total power and it can put its geothermal developments in places people don't care so much about.
Every geothermal plant I've seen, in person and in pictures, covers a lot of land and is ugly. There's no way around it. It takes a lot of wells and a lot of pipe, which in turn means a lot of roads, to collect the steam or hot water. The plants themselves are very large relative to the amount of power generated, because they're starting with a low-grade heat source that is sometimes barely warmer than the "waste heat" from a nuclear plant and is contaminated with minerals that would ruin turbines if used directly. Often the "used" water must be handled as hazardous waste due to natural toxic minerals in it, which usually means pumping the spent fluid back into the ground. That in turn takes more energy and machinery.
The idea that you just drill a well, get live steam out of it, and feed it to a turbine is a pure fantasy. Geothermal power can definitely be made to work, but it's bulky and ugly and has its own set of environmental impacts. And yes it DOES generally dry up nearby natural hot springs.
On the balance, the Japanese may decide that it's worth it. They may prefer the known, moderate, environmental damage from geothermal plants over the very unlikely but potentially catastrophic environmental damage from nuclear plants. It's a complex decision for intelligent people to make, and the last thing they need is a "tough and wise government" to pooh-pooh their very legitimate fears of environmental damage from geothermal development, just as a "tough and wise government" once pooh-poohed their fears of nuclear power.
I might suggest a radical third course. Japan is an aging country. It no longer is, or needs to be, a manufacturing powerhouse. Japanese companies can have their manufacturing done elsewhere in the world. Japan doesn't need its own heavy industry any more. With tight immigration restrictions, its population will likely begin a long-term decline. Rather than trying to maintain a power supply equal to what it had during its industrial and population growth phase, why not accept that the country is basically retiring and doesn't need the terrawatts it once needed? Fire up the remaining nuclear plants that are judged safe, and as they wear out, shut them down and don't replace them with anything. Let the nation's power supply decline to track its declining reliance on heavy industry and its declining population. There's no law of nature that says every country has to always grow.

Why is red shown in the key for "copper", and yet it appears nowhere on the graphs?