I don't have any numbers to back this up, but the way I've heard it end of life care cost is much higher in the US than in Europe or elsewhere. Not just absolute but also relative to total health care cost. You're wrong.
The amount of people in the US that state they want doctors to do everything they can to stay alive a few more days is staggering. Even if it will make them miserable, and success of treatment is measured in adding days to the cancer patient's life, by slowing the progression (while getting the patient more and more sick), not in how it cures them of anything.
Ask any cancer doctor: there is a rather large very religious group of US residents that doesn't care about suffering and just does not want to die. So millions are wasted to add suffering and paradoxically postpone their trip to heaven, wasting money that can save other people's live in the process.
I believe, like many others have stated, that the time to pull the plug on a terminally ill patient is when they don't want life anymore. Now this won't affect cost here in the US much unless that terribly large religious subset (please, I'm not talking all religious people, don't get me wrong) stops fearing death so much. http://en.wikipedia.org/wiki/Terror_management_theory

Indeed, the oldstable (etch) version of debian, which is still supported wrt security has floppy images.

For example, these can be found here: http://ftp.nl.debian.org/debian/dists/etch/main/installer-i386/current/images/floppy/

It should be easy to stay under 700 Mb, even with x. Just don't install gnome/kde , but do xfce or lxde instead.
you can even get to lenny (current stable version) by changing /etc/apt/sources and apt-get dist-upgrade.

All this talk about legality becomes quite ironic when you consider Philips is actually a Dutch company where pot is semi-legal.....

Glad they set the cut-off at 40 then, honestly there is nobody in the world with a bmi that high who is not extremely fat.
There are no borderline cases anymore when you get beyond 35.
Bmi is a bad measure for individual cases, because healthy muscularly people might fall into the wrong category, but when we are talking a population and we are talking a bmi of 40 there is no discussion.

But it can't. The test does most likely not correlate with actual behavior at all.

To just say it bluntly: almost all personality tests are completely unreliable, the best ones are at most somewhat reliable. This means that most people, when they do the same personality test some time later, score differently. This is a terrible problem, as the point of these tests often is to measure a personality, something presumably stable over time. If a test used to measure a trait gives unreliable test results, it apparently can not be measuring a stable trait, and what it does measure can not be used as a predictor of behavior! Even the traits as measured in the Myers-Briggs questionnaire (the most used test in this country to measure personality) have terrible issues with reliability (for an easy quick link: http://en.wikipedia.org/wiki/Myers-briggs#Reliability). A lack of reliability in a measure is extremely problematic when the measure is used for predictions, as lacking reliability literally means lacking predictive power.

Now this poor reliability I mentioned is the test-retest reliability, the reliability for a test to predict itself in the future. that being bad is made even worse because more often than not these tests are used to predict actual behavior. And the reliability of any pen-and-paper test to predict actual behavior is even worse than the retest reliability.

This problem of low reliability is the biggest issue and concerns all personality scores ever made. However, most tests, unlike the big five or the Myers-Briggs tests, suffer from horrible validity issues as well. Any ad hoc test created by employers is probably not even measuring what it's intended to measure due to it's creators not doing a proper factor analysis and such to test for underlying constructs, etc. (see http://en.wikipedia.org/wiki/Validity_(statistics) )

So besides being non-predictive of anything, these tests are not even measuring anything useful if they were.

Could you clarify this distinction you make between the passive aggression of listening to pantera and active aggression that video games 'are'?
Does this mean passively watching violent movies is also passive aggression? And killing a mosquito is active aggression?
It sounds more like personal preference to me, clad in nice sounding terms.

It's been shown in many sound social psychological studies over decades that in children there is a strong correlation between watching violent tv and aggressive behavior, between playing violent video-games and aggressive behavior and between listening to aggressive music and aggressive behavior.
Go google (google scolar) yourself or look it up on wikipedia.

--There has never been any study proving a causal relationship between these (with behavior being the dependent factor) where the effect lasts for more than a few minutes. --

The catharsis theory ("I go to martial arts school so I don't have to be violent at home", "I listen to pantera once I'm at home so I can be more calm when I'm at work") is a Freudian theory disproved ages ago as well. I'm sure people can peruse the relevant social and personality psychology literature themselves on this. (journal of personality and social psychology, etc. )