Half inch mild steel might stop lead. Even .308 AP will go straight through without even slowing down. I have a whole box of such plates in my shop.

Wake me when it meets MILSTD 810-G.

I know exactly what you mean, metlin. I've been doing business case analysis for a startup I'm developing which will produce a tangible product with a high profit margin. A single machine in the production process has 6 zeroes in the price tag and the amount of money needed by the company over 5 years has seven zeroes. While I was able to fund the basic research with my own checkbook, scaling it into a profitable product with that funding source is impossible.

That failure pattern is called the bathtub curve or Weibull distribution.

Honestly, some of the most interesting and intelligent people I have ever met were art historians. On the other hand, it isn't a degree you get as professional training to do something lucrative.

The XFS guys are right. I've lost data multiple times on XFS due to their disk caches being enabled in Suse 11.4. My disks are ST9750420AS using bios raid. I finally had to disable disk write caches on bootup. These losses were not even due to power failure: these losses were incurred during graceful system shutdown.

A good use for stupidfilter http://stupidfilter.org/ perhaps?

Untar in this context simplym means heating up the sundial with a giant magnifying glass and hoping the tar runs off . .

It wasn't what they said but the aspersions in the way they said it that they made listening to them interesting. . .

I do chemical engineering research in my spare time. I actually had to use calculus a few weeks ago to derive an equation for making a mixture of 2 solvents as similar as possible to the properties of a third solvent. For two solvents, the answer was a calculus gimme.

For three solvents, I had to break out multivariable calculus and Lagrange multipliers which didn't generate a solution but instead reduced the problem to a 4x4 matrix inversion. The inversion would have been most easily solved by feeding the matrix into the LAPACK library but a pedant could have solved it in his own code using Cramer's rule.

In my case, calculus wasn't enough, it only reduced the problem to linear algebra. Basically, you have to have taken enough math to recognize when you have transformed a problem form an intractable form to a tractable form.

Actual scientists are the most likely to mess these kinds of stuff up. They often weren't trained in the nuances of calculation. I actually head a bunch of physicists at Los Alamos arguing about whether computer science was a science!

The clinicians using it may have been honest and competent but the people designing it were anything but.

Trying to mod insightful and accidentally modded redundant. Posting to cancel

I believe that it's a lawyer and not a developer that needs to be consulted in this case as there are legal precedents. Just try googling for uniform commercial code software development

I'm blowing a couple Mod points but. . . Take a look at www.brickwall.com for surge suppressors. These are not whole house units but they are not damaged by 1000 consecutive surges at the maximum energy at which IEEE tests. They also have panel mounted ones that can be used to protect a whole circuit. They are basically an analog lowpass filter on steroids and as a result suffer no damage from surges that would completely destroy a MOV based surge protector. I've got one on my stereo. I've never looked at the results with an o-scope or anything but the engineering principle on which they are based is sound.