The insulating materials in my house do a great job of stopping EM fields around 600 Terahertz.
The sheet of white paper on my desk is a good EM reflector throughout the visible spectrum, but doesn't conduct worth a damn. I can shine a 5 Watt green laser on it and it doesn't even heat up noticeably.
And so on...
Even without metal, there are numerous cases of first- and second-degree burns to the wrist area, which are presumed to be due to the arms crossed and over the chest acting as a pickup loop, but the most severe burns seem to arise from resonance between the RF drive and metal loops in contact with the patient (implantable devices or monitoring wires, such as pulse oxymeters and ECG leads).
Chernobyl is not the only nuclear accident in FSU. It's the only reactor breach, but there have been severe accidents in fuel processing and waste storage sites, such as the one in Tomsk April 6 1993.
An agent can be tremendously useful in steering a manuscript to people who will be able to evaluate it well. A good agent will have a lot of knowledge you won't find on slashdot about getting the MS to the right readers.
For a good time, try some Stanislaw Lem. Solaris is big because of the movie, but all hackers should really read the Cyberiad. Memoirs Found in a Bathtub is also first rate.
Philip K. Dick is also well worthwhile if you haven't tried him. I like Do Androids Dream of Electric Sheep and A Scanner Darkly
When you're done with Dick's mindfucks on consciousness and identity, try some other things than SF. I would particularly recommend Orhan Pamuk's My Name Is Red---a postmodernist mystery set in 16th century Istanbul (as the title suggests, one chapter is narrated by a particular shade of red)---and The White Castle (it's not about tiny hamburgers), or Nabokov's Pale Fire.
There are also some fantastic novels about doing science in the real world, as opposed to the more fantastical stuff that gets called SF. The best of the crop is Carl Djerassi's Cantor's Dilemma, which is perhaps the best novel of science ever.
Back to the predictable, if you really want things only about sf-geek-hackerdom, pick up Vernor Vinge's "True Names" and John Brunners' Shockwave Rider, which was perhaps the first cyberpunk book of all.
When was the last time a coal powerplant had a catastrophic failure that endangered all who lived near it?
Most fatalities from coal are not from power-plant accidents but from mining. Mining accidents mostly kill miners (who cares about them?), but also can kill many people who live near the mine. The 1972 flood at the Buffalo Creek Coal Mine in West Virginia killed 125 people living nearby, injured over 1000, and completely destroyed 500 homes.
Worldwide, tens of thousands of deaths per year occur from coal-mining accidents, and that doesn't count slow deaths from black-lung and other chronic conditions that afflict miners. In India, the death rate is equivalent to one Bhopal per month. In China, around 5000 people per year are killed in coal mining accidents.
Compare all this to the estimated 2500 deaths due to Chernobyl.
In his autobiography, Ansel Adams describes how he told an interviewer that he would like to "drown Ronald Reagan in my martini," which led to an invitation to lunch at the White House. The only thing the two agreed on about the environment was their support of nuclear power. Those who don't like books can read a paraphrased account here.
Not to mention that if anyone has a claim to sue Linus, it's the people who created Minix, for creating a workalike - and even then, he didn't copy code. Go look at the UNIX heritage charts for a much better understanding.
You don't have to copy anything to infringe on a patent. I can infringe on the Unisys LZW patent by writing my own LZW code. All I have to do is use (whether by copying or by innocently reinventing) Terry Welch's algorithm during the next month in the USA.
Similarly, if Linus introduced into the kernel either his own code or code donated by someone else (IBM) that implements any algorithms for which SCO holds patents, even if the code in question comprises completely original implementations of those algorithms, then by redistributing the kernel without a license from SCO, Linus is infringing on SCO's patents.
Never trust a computer proffesional that doesnt list computer as a hobby.
In Soul of A New Machine, Tracy Kidder reports that Data General got the best productivity from its engineers by hiring only applicants who didn't list computer as a hobby.
Since the shift requires shooting the photonic crystal with a bulled, you'd have a room full of drunk horny single people shooting guns. Darwin award, anyone?.
No. My argument is to keep flying shuttles and keep killing astronauts at a rate of about one crew per decade. That would leave a lot more money for research. As for astronauts, they have known the risks all along. NASA and the National Academy of Science have reviewed and published the probabilistic risk assessments all along and these numbers have not scared astronauts away.
After the last shuttle crashes, we could ask whether the research conducted on the shuttle was more productive in saving lives per dollar spent than terrestrial research or that conducted on unmanned spacecraft and decide on that basis whether to build another fleet.
The EPA has announced that they find it worth no more than $3.7 million dollars per life saved to regulate pollution.
NASA will spend of order $100 million per life saved to make the shuttle safer. Are astronauts really worth so much more than the rest of us?
Meanwhile, nobody seems to have noticed that the rate of shuttle losses is completely statistically consistent with what NASA has been telling us all along:
According to NASA, the probability of a critical failure was 1 in 145 before 1998 and 1 in 245 from 1998 on. Plugging these numbers into the binomial probability distribution, we find that the probability of 2 shuttle losses by 2003 is about 16%. Most people find 5% to be the threshold for statistical significance, so there is no good reason to believe that NASA got the risks wrong based on the two critical failures!
This is not rocket science, it's freshman statistics.
A significant part of the environmental cost of computers is expended in manufacturing the computer, before you even buy it. Semiconductor and PC board manufacturing use tremendous quantities of fresh water (about ten gallons per chip and a total of 8,000 gallons per computer), which has serious environmental consequences in the American West and in many parts of the third world. Of course, as long as the state of California subsidizes its rice farmers' water, there are more important places to complain about this.
Also, semiconductor manufacturing uses lots of quite nasty chemicals and while the organics can be incinerated, the heavy metals are difficult to dispose of safely for the long term and there is always the inevitable discharge of toxic pollutants into the air or water surrounding the factory.
Finally, both manufacturing and operating computers use lots of electricity, which is usually generated by plants that produce lots of greenhouse gases.
Besides worrying about recycling, you also want to worry about all these environmental costs.
Seems to me that National Instruments could use this recommendation to develop a web services based version of LabView
They already pretty much have this, only they call it "DataSocket." Some folks at MIT have written a tutorial on how to put data on your web page using a Java DataSocket client.
How about error recovery and integrity checking? There are no checksums or any kind in CVS repositories, so you can never find out whether a repository file has been corrupted until you try to check out an old revision or perform a branch merge and find out that there's a broken delta in the way.
To me, the lack of good integrity checking and error recovery is the greatest weakness of CVS. It means that you can't rely on recent backups, but must keep every CD you've ever burned of your repository.
When I was learning about molecular spectroscopy, my advisor handed me a copy of J. Steinfeld's Molecules and Radiation and told me, "The best way to learn about molecular spectroscopy is to read this book and find all the errors." Indeed, knowing the book was chock full of errors, I read it more carefully than I did most textbooks, rederived all the equations myself, and learned a lot.
You don't have to get college textbooks at several hundred dollars a pop. You can buy really good maths and science textbooks in paperback from Dover for around ten to twenty dollars. What you get is generally a classic originally published 20-50 years ago. As was common then, these books have informative content instead of color illustrations and cute sidebar articles. They also don't have the latest developments, but it's very likely that what you want to learn was thoroughly developed and well-understood back then.
These are books that have stood the test of time and are generally well written, well edited, and quite accurate.
Sourceforge has some serious problems with respect to managing CVS archives. Most important is the dangling lock problem. If a transaction aborts and leaves a dangling lock file, nobody, not even the project admin, can even get read access to cvs until someone emails a sourceforge admin and gets them to go delete the lock file.
This kind of thing might be fine for small projects, but I don't think it's acceptable for a large project.
Also, when I take off my developer hat and put on my consumer hat, when I try to search for projects using Sourceforge's search page, the search engine is disabled over 90% of the time. I would not call that industrial strength.
How many people would need to reviewhttp://www.riaa.org as porn before they'd rate it as such? Thereafter, one could start asking them to block home pages for legislators who voted for DMCA, Microsoft, whoever we don't like. If there were enough traffic, perhaps the slashdot effect be a powerful lobby for a nice little blacklist.
OK. I think of libertarians as liberals and people like George W. Bush, Rick Santorum, and Trent Lott as conservatives. Now you've cleared up my misunderstanding. Historically, liberals were those, such as John Locke and Adam Smith, who embraced enlightenment thinking and wanted to base government and laws on liberty and rationality, where conservatives (Tories) wanted to retain God-given natural law, monarchy, hereditary station in life, etc. This translates today into liberals who want to get the government and laws out of our bedrooms and away from our liberties, and conservatives who want government and laws to enforce "traditional Judeo-Christian values."
in the U.S. they have two parties, just as we [in England] have two parties. They have the Republican party, which is like our Conservative party. And they have the Democratic party, which is like our Conservative party.
But now that I understand where you're coming from, we don't need to split hairs over political labels.
I'm not sure what to make of your statement that you're concerned more with criminal than civil litigation. This whole thread was a about civil matters (California's laws on implied warranties). As to criminal matters, local judges and juries in the South in the 1960s accurately reflected racist community values and exonerated some awful murderers and terrorists, some of whom took advantage of the Constitutional protection against double jeopardy and sold the stories of their brutal acts to the press.
If you think this is all ancient history, look at what local judge Edward Self did in Tulia Texas. When rogue cop Tom Coleman framed about 15% of the black population of Tulia Texas for selling cocaine in 1999 (one 57 year old hog farmer was sentenced to 99 years), Judge Self refused to admit evidence introduced by defense lawyers that demonstrated a pattern of deceit and shoddy police work by Detective Coleman. Later Judge Self lied about his refusal to admit this evidence and despite being caught lying and forced to recuse himself from appeals of the Tulia cases, Judge Self was re-elected. This spring, the cases were re-opened by higher authorities who don't have to stand for election in Tulia. Detective Coleman has been charged with three counts of aggravated perjury and all the convictions are being vacated. The State of Texas is preparing to pay the victims up to $3000 for each year they wrongfully spent in prison.
Before you get too enamored of direct local democracy, I would recommend re-reading Federalist X on the danger of faction and the tyranny of the majority. This spring, people farther removed from the local level got involved and
Well, given that I'm quite a conservative thinker, I would say that your charactarization is suspect. I would say that *centrists* are the ones that would prefer slow change and predictability.
conservative ADJECTIVE: 1. Favoring traditional views and values; tending to oppose change. 2. Traditional or restrained in style: a conservative dark suit. 3. Moderate; cautious: a conservative estimate. 4a. Of or relating to the political philosophy of conservatism. b. Belonging to a conservative party, group, or movement. 5. Conservative Of or belonging to the Conservative Party in the United Kingdom or the Progressive Conservative Party in Canada. 6. Conservative Of or adhering to Conservative Judaism. 7. Tending to conserve; preservative: the conservative use of natural resources.
conservatism NOUN: 1. The inclination, especially in politics, to maintain the existing or traditional order. 2. A political philosophy or attitude emphasizing respect for traditional institutions, distrust of government activism, and opposition to sudden change in the established order. 3. Conservatism The principles and policies of the Conservative Party in the United Kingdom or of the Progressive Conservative Party in Canada. 4. Caution or moderation, as in behavior or outlook.
You may want to use the word in different ways than the dictionary (hell, nobody in the press uses either liberal or conservative correctly, so why should you?), but I stand by my characterization.
Thanks for the clarification of your political leanings. I am confused about your comments on precedent, though. You write, "Case law better serves situations in Federal and State Court where there is a much greater interest in consitancy." There are only two kinds of courts: state and federal. There is no separate "local" court. Perhaps you're thinking of county or district divisions of state court.
As to the benefit of local judges, you may want to think again. A local judge in the Bronx recently presided over a lawsuit in which a local jury awarded $51 million to Darryl Barnes, who was paralyzed when he drew a TEC-9 on a police officer who fired first. Bronx voters favor judges who dispense Robin Hood justice.
Similarly, local courts have made Mississippi and many other southern states "judicial hellholes" for product liability litigation according to the American Tort Reform Association. The problem here is that the voters in the court's jurisdiction tend to identify with the plaintiff. The corporations sued generally have their factories elsewhere, so it doesn't affect the local voters if jobs dry up. On the other hand, a large award can bring a lot of money to a small county. It is this abuse of discretion by local judges and juries that makes the U.S. Chamber of Commerce call for moving litigation out of local courts and into federal courts while consumer rights groups want to keep litigation in local courts.
As to your comment, "If someone has really done something so wrong that they deserve to pay out 30 billion dollars in punative damages then they should go to jail instead," I would only direct your attention to O.J. Simpson. O.J. did not go to jail (again, your beloved local justice did a fine job with him, as it did with the murderers of Emmett Till) and punitive damages were the most justice he received.
At the same time, the number of liability suits ending in bizarre awards is much smaller than most people think. Just as the press tend to exaggerate spectacular events, such as civilian casualties in Iraq, and make them seem more common than they are, they also exaggerate the frequency of liability blunders (absurd verdicts, excessive awards, etc.) and do not follow up six months later when absurd awards are thrown out on appeal.
0.02% of all civil cases handled by the state courts concern product liability.
The defendants win more than half of these cases.
When plaintiffs do win product liability cases, more than half the awards are less than $27,000.
Awards over $1 million are most common when the plaintiff has suffered grievous injury (paralysis, brain damage, amputation) and over half of these large awards are either reversed or reduced substantially by the trial judge or on appeal.
Only a small fraction of a percent of all findings for the plaintiff award punitive damages. Between 1965 and 1994, there were 379 punitive damage awards in the U.S. in product liability lawsuits. Half of these awards were less than $50,000.
According to the Consumer Federation of America, liability suits add about 0.26% to the cost of consumer goods. This number is similar to what the National Association of Insurance Commissioners found.
It's a delicate balance between the inflexibility of precedent and the caprice of the individual judge. Many, perhaps most, businesses would prefer to work in an unfair but predictable legal arena than an unpredictable one, even if the latter were somewhat fairer on average. Even when you disagree with the prevailing standards, you can at least make coherent plans for risk management. When individual judges have too much discretion, liability suits become too unpredictable to plan for. Managing an uncertain risk can easily cost more than managing a greater but more consistent risk because what kills you in financial risks are the tails of the distribution (That's why Long Term Capital Management averaged a positive return on investment in 1998-1999, but went bankrupt in 1998 because of very improbable fluctuations.)
Conservative thinkers tend to prefer slow change and predictability, which is why they push for mandatory sentencing laws, caps on civil awards, and other measures to limit judicial discretion. Since this nation is strongly conservative at the moment, you may have to wait some time before liberals return to power and push for greater judicial independence.
How do you see this applying to software product liabilities?
There have been cases where software bugs in medical equipment killed people. In those cases, there would be strong precedent for product liability lawsuits.
Slashdot Mirror
The sheet of white paper on my desk is a good EM reflector throughout the visible spectrum, but doesn't conduct worth a damn. I can shine a 5 Watt green laser on it and it doesn't even heat up noticeably.
- Dempsey MF, Condon B., "Thermal injuries associated with MRI," Clin Radiol 2001 Jun;56(6):457-65
- Brown TR, Goldstein B, Little J., "Severe burns resulting from magnetic resonance imaging with cardiopulmonary monitoring. Risks and relevant safety precautions," Am J Phys Med Rehabil 1993 Jun;72(3):166-7.
- Schaefer DJ, "Safety aspects of radiofrequency power deposition in magnetic resonance," Magn Reson Imaging Clin N Am 1998 Nov;6(4):775-89.
And so on... Even without metal, there are numerous cases of first- and second-degree burns to the wrist area, which are presumed to be due to the arms crossed and over the chest acting as a pickup loop, but the most severe burns seem to arise from resonance between the RF drive and metal loops in contact with the patient (implantable devices or monitoring wires, such as pulse oxymeters and ECG leads).Chernobyl is not the only nuclear accident in FSU. It's the only reactor breach, but there have been severe accidents in fuel processing and waste storage sites, such as the one in Tomsk April 6 1993.
An agent can be tremendously useful in steering a manuscript to people who will be able to evaluate it well. A good agent will have a lot of knowledge you won't find on slashdot about getting the MS to the right readers.
Philip K. Dick is also well worthwhile if you haven't tried him. I like Do Androids Dream of Electric Sheep and A Scanner Darkly
When you're done with Dick's mindfucks on consciousness and identity, try some other things than SF. I would particularly recommend Orhan Pamuk's My Name Is Red---a postmodernist mystery set in 16th century Istanbul (as the title suggests, one chapter is narrated by a particular shade of red)---and The White Castle (it's not about tiny hamburgers), or Nabokov's Pale Fire.
There are also some fantastic novels about doing science in the real world, as opposed to the more fantastical stuff that gets called SF. The best of the crop is Carl Djerassi's Cantor's Dilemma, which is perhaps the best novel of science ever.
Back to the predictable, if you really want things only about sf-geek-hackerdom, pick up Vernor Vinge's "True Names" and John Brunners' Shockwave Rider, which was perhaps the first cyberpunk book of all.
The heck with NIMBY. My back yard's too small. I say, Build Absolutely Nothing Anywhere Near Anything.
Most fatalities from coal are not from power-plant accidents but from mining. Mining accidents mostly kill miners (who cares about them?), but also can kill many people who live near the mine. The 1972 flood at the Buffalo Creek Coal Mine in West Virginia killed 125 people living nearby, injured over 1000, and completely destroyed 500 homes.
Worldwide, tens of thousands of deaths per year occur from coal-mining accidents, and that doesn't count slow deaths from black-lung and other chronic conditions that afflict miners. In India, the death rate is equivalent to one Bhopal per month. In China, around 5000 people per year are killed in coal mining accidents.
Compare all this to the estimated 2500 deaths due to Chernobyl.
In his autobiography, Ansel Adams describes how he told an interviewer that he would like to "drown Ronald Reagan in my martini," which led to an invitation to lunch at the White House. The only thing the two agreed on about the environment was their support of nuclear power. Those who don't like books can read a paraphrased account here.
You don't have to copy anything to infringe on a patent. I can infringe on the Unisys LZW patent by writing my own LZW code. All I have to do is use (whether by copying or by innocently reinventing) Terry Welch's algorithm during the next month in the USA.
Similarly, if Linus introduced into the kernel either his own code or code donated by someone else (IBM) that implements any algorithms for which SCO holds patents, even if the code in question comprises completely original implementations of those algorithms, then by redistributing the kernel without a license from SCO, Linus is infringing on SCO's patents.
In Soul of A New Machine, Tracy Kidder reports that Data General got the best productivity from its engineers by hiring only applicants who didn't list computer as a hobby.
Since the shift requires shooting the photonic crystal with a bulled, you'd have a room full of drunk horny single people shooting guns. Darwin award, anyone?.
After the last shuttle crashes, we could ask whether the research conducted on the shuttle was more productive in saving lives per dollar spent than terrestrial research or that conducted on unmanned spacecraft and decide on that basis whether to build another fleet.
NASA will spend of order $100 million per life saved to make the shuttle safer. Are astronauts really worth so much more than the rest of us?
Meanwhile, nobody seems to have noticed that the rate of shuttle losses is completely statistically consistent with what NASA has been telling us all along:
According to NASA, the probability of a critical failure was 1 in 145 before 1998 and 1 in 245 from 1998 on. Plugging these numbers into the binomial probability distribution, we find that the probability of 2 shuttle losses by 2003 is about 16%. Most people find 5% to be the threshold for statistical significance, so there is no good reason to believe that NASA got the risks wrong based on the two critical failures!
This is not rocket science, it's freshman statistics.
Also, semiconductor manufacturing uses lots of quite nasty chemicals and while the organics can be incinerated, the heavy metals are difficult to dispose of safely for the long term and there is always the inevitable discharge of toxic pollutants into the air or water surrounding the factory.
Finally, both manufacturing and operating computers use lots of electricity, which is usually generated by plants that produce lots of greenhouse gases.
Besides worrying about recycling, you also want to worry about all these environmental costs.
They already pretty much have this, only they call it "DataSocket." Some folks at MIT have written a tutorial on how to put data on your web page using a Java DataSocket client.
Someone else has implemented a Python DSTP server and client, so you don't have to buy from NI.
To me, the lack of good integrity checking and error recovery is the greatest weakness of CVS. It means that you can't rely on recent backups, but must keep every CD you've ever burned of your repository.
When I was learning about molecular spectroscopy, my advisor handed me a copy of J. Steinfeld's Molecules and Radiation and told me, "The best way to learn about molecular spectroscopy is to read this book and find all the errors." Indeed, knowing the book was chock full of errors, I read it more carefully than I did most textbooks, rederived all the equations myself, and learned a lot.
You don't have to get college textbooks at several hundred dollars a pop. You can buy really good maths and science textbooks in paperback from Dover for around ten to twenty dollars. What you get is generally a classic originally published 20-50 years ago. As was common then, these books have informative content instead of color illustrations and cute sidebar articles. They also don't have the latest developments, but it's very likely that what you want to learn was thoroughly developed and well-understood back then. These are books that have stood the test of time and are generally well written, well edited, and quite accurate.
This kind of thing might be fine for small projects, but I don't think it's acceptable for a large project.
Also, when I take off my developer hat and put on my consumer hat, when I try to search for projects using Sourceforge's search page, the search engine is disabled over 90% of the time. I would not call that industrial strength.
How many people would need to review http://www.riaa.org as porn before they'd rate it as such? Thereafter, one could start asking them to block home pages for legislators who voted for DMCA, Microsoft, whoever we don't like. If there were enough traffic, perhaps the slashdot effect be a powerful lobby for a nice little blacklist.
As Jonathan Miller once said,
But now that I understand where you're coming from, we don't need to split hairs over political labels.I'm not sure what to make of your statement that you're concerned more with criminal than civil litigation. This whole thread was a about civil matters (California's laws on implied warranties). As to criminal matters, local judges and juries in the South in the 1960s accurately reflected racist community values and exonerated some awful murderers and terrorists, some of whom took advantage of the Constitutional protection against double jeopardy and sold the stories of their brutal acts to the press.
If you think this is all ancient history, look at what local judge Edward Self did in Tulia Texas. When rogue cop Tom Coleman framed about 15% of the black population of Tulia Texas for selling cocaine in 1999 (one 57 year old hog farmer was sentenced to 99 years), Judge Self refused to admit evidence introduced by defense lawyers that demonstrated a pattern of deceit and shoddy police work by Detective Coleman. Later Judge Self lied about his refusal to admit this evidence and despite being caught lying and forced to recuse himself from appeals of the Tulia cases, Judge Self was re-elected. This spring, the cases were re-opened by higher authorities who don't have to stand for election in Tulia. Detective Coleman has been charged with three counts of aggravated perjury and all the convictions are being vacated. The State of Texas is preparing to pay the victims up to $3000 for each year they wrongfully spent in prison.
Before you get too enamored of direct local democracy, I would recommend re-reading Federalist X on the danger of faction and the tyranny of the majority. This spring, people farther removed from the local level got involved and
According to The American Heritage Dictionary,
You may want to use the word in different ways than the dictionary (hell, nobody in the press uses either liberal or conservative correctly, so why should you?), but I stand by my characterization.As to the benefit of local judges, you may want to think again. A local judge in the Bronx recently presided over a lawsuit in which a local jury awarded $51 million to Darryl Barnes, who was paralyzed when he drew a TEC-9 on a police officer who fired first. Bronx voters favor judges who dispense Robin Hood justice.
Similarly, local courts have made Mississippi and many other southern states "judicial hellholes" for product liability litigation according to the American Tort Reform Association. The problem here is that the voters in the court's jurisdiction tend to identify with the plaintiff. The corporations sued generally have their factories elsewhere, so it doesn't affect the local voters if jobs dry up. On the other hand, a large award can bring a lot of money to a small county. It is this abuse of discretion by local judges and juries that makes the U.S. Chamber of Commerce call for moving litigation out of local courts and into federal courts while consumer rights groups want to keep litigation in local courts.
As to your comment, "If someone has really done something so wrong that they deserve to pay out 30 billion dollars in punative damages then they should go to jail instead," I would only direct your attention to O.J. Simpson. O.J. did not go to jail (again, your beloved local justice did a fine job with him, as it did with the murderers of Emmett Till) and punitive damages were the most justice he received.
At the same time, the number of liability suits ending in bizarre awards is much smaller than most people think. Just as the press tend to exaggerate spectacular events, such as civilian casualties in Iraq, and make them seem more common than they are, they also exaggerate the frequency of liability blunders (absurd verdicts, excessive awards, etc.) and do not follow up six months later when absurd awards are thrown out on appeal.
Here are some facts, courtesy of the Center for Justice and Democracy and Public Citizen:
Conservative thinkers tend to prefer slow change and predictability, which is why they push for mandatory sentencing laws, caps on civil awards, and other measures to limit judicial discretion. Since this nation is strongly conservative at the moment, you may have to wait some time before liberals return to power and push for greater judicial independence.
There have been cases where software bugs in medical equipment killed people. In those cases, there would be strong precedent for product liability lawsuits.
Suppliers to the military are harder to sue, which is probably good news to the folks whose bugs killed soldiers when their mortar targeting software made incorrect assumptions about target altitude or when a Patriot missile targeting system's clock overflowed after 8 hours.
For further reading on software liability issues, see this Business Week article, which was discussed on /. and badsoftware.com, which surveys software liability issues from a consumer's perspective.