http://docs.oracle.com/cd/E19957-01/806-3568/ncg_goldberg.html "What Every Computer Scientist Should Know About Floating-Point Arithmetic", by David Goldberg, published in the March, 1991 issue of Computing Surveys

I don't know why anyone thought this was surprising (it would have been surprising if they didn't get different results, given that some use GPUs, some don't, etc.). What does tend to get "amusing" is that even with the same processor folks get different results (sometimes due to software issues, chip rev issues, or actual hardware bugs that go undetected ... but are minor enough to remain so unless someone gets really careful and whips out the old logic analyzer).

To get a ruling on whether you may do what you want. Otherwise, as others have noted, you may be very deep waters (not only will you be in violation, but anyone in the organization using the service will be, and you will have induced them to do it. Think serious civil as well as criminal consequences).

From a technology angle, it may be "possible" if the folks in charge sign off.

"All" you need to do is encrypt the data before it goes offsite, encrypt it well enough that the data is protected commensurate with its value, etc.

For commercial users, https://jungledisk.com/ provides a very usable interface and GUI. Of course, if the client isn't trustworthy (and you have to take their word for it ;>) that goes out the window even if the algorithms are secure themselves ;>

I use it for some SOHO confidential data; it wouldn't be the end of the world if the data were disclosed, but we have committed to make good faith effort(s) to keep it secure, so we do (rather than moving files to subs via email, etc.). Not all subcontractors could handle sftp and friends.

Yes, as many have noted.... client organization unhappiness and excess spend are objective metrics which call into question what's going on.

But ... dig deeper.... are the client organizations behaving rationally or are they constantly asking for change after change? Obviously the organization is dysfunctional ... but is it the IT director, the entire IT department or the organization as a whole?

If it IS the IT director... I imagine the issue he's(or she) has been there for a long time; HR will want "proof" that replacing them isn't age discrimination or any such thing. If the client organizations are healthy and have reasonable expectations and your impression of the "line worker" IT folks is good, perhaps you need to have a heart to heart with the IT director. If he is "parroting" what his staff tell him, he's made some poor hires. He may even understand that, but lack clues as to how to hire better. See if they are rational, and self-aware enough to recognize their limitations and work with he/she/it to hire someone to be the "technical honcho" ... chances are the IT director wouldn't have gotten the job and kept it so long if they didn't have good relationships with the executive staff. That DOES have value. Leverage it, and help them improve ... if they are capable of it.

Just sacking someone is seldom enough; they will have built up an empire of mixed wood (some dead, some living, some actually thriving) and you need to help them prune and fertilize ... not just toss the whole tree away (unless, of course, their entire IT department really could be replaced with SaaS and IaaS and have done with it).

The details of the organization and people matter.

As earlier posters noted, if this is all news to you, you might not be the right consultant.

By not getting through to you ;>

As Blackstone put it, and our Founding "Fathers" enshrined into actionable text, "It is better that ten guilty persons escape than that one innocent suffer". The entire US legal system is based on this notion. It is hard for the State to prove Guilt (beyond a reasonable doubt, hands tied in various ways, etc.). Yes, you could have probably developed a vast complex of rules that could exist without the 5th amendment and still accomplish its ends, but unlike the legislature of today, back then they sought brevity and relative simplicity of rules.

Even with all the rules stacked in favor of the defendant, innocents get convicted.

Imagine that there was not 5th amendment, then the likelihood of government focus on getting "lie detectors" ruled legitimate in court. Since all they actually measure is stress, they'd pepper the suspect with questions and convict of things they were nervous about (objective evidence no longer required). Just one of many probable outcomes of removing this protection.

That your gripe is with differential equations suggests that your department is just using the Math courses as a screening process (eliminate the chaff). Statistics (parametric and non-parametric), linear algebra, number theory, numerical methods ... these all have direct application to real world problems folks face in applying computers to business or engineering problems.

Number theory is the basis for cryptosystems. You probably won't be developing your own, but understanding a bit of why they work (or don't) is an example of how advanced mathematics impacts our day to day life in CS applications.

Even if you don't become a "data scientist" understanding statistics (correct and incorrect usage) are key to performance analysis, system tuning, etc.

Numerical methods help one appreciate entire classes of errors which computers make by design; not critical for an OS developer (no fp in kernels) but someday, somehow you may find yourself dealing with floating point computations ... learning about the fine points (see, for example, the Goldberg paper "What Every Computer Scientist Should Know about Floating Point Arithmetic" http://docs.oracle.com/cd/E19957-01/806-3568/ncg_goldberg.html).

The list goes on. But differential equations really shouldn't be part of it. Pity that few CS departments work closely with Math departments to craft courses where the subject matter not only matters, but the linkage is made explicit.

The 2002 version of the standard added object features. While not my first choice of languages, it is typically not cheaper nor safer to rewrite large amounts of working tested code. Yes, you might do better with a clean sheet of paper and a decade or so, but most IT organizations don't have that luxury.

My favorite COBOL nerdy feature died many versions of the Standard ago (MOVE CORRESPONDING). It was my favorite not because it was a terrific feature, but it was just so unique to COBOL.

Cloud computing is, as a business model, a return to mainframe timesharing services such as dominated in the original COBOL and PL/I eras. It really is not a stretch to see IBM update their zSeries environment to easily enable leveraging the COBOL code base.

Yes, you can (and more cheaply per IBM MIP) run Linux on your zSeries hardware, so you can mix and match (write new applications, or layers in newer environments) ... but there is no need to toss out dull boring functional code that just happens to be business critical.

No doubt the sufficiently intrepid IT staffer could rewrite all the COBOL in Haskell or Perl .. (or for extra credit in REXX) but would it really be an improvement? Indeed, just validating that the new code is logically equivalent to the original code for ALL input sets would be a huge investment ... never underestimate the cost (or importance) of Test and Validation.

The folks without MDs and who can't prescribe drugs are asserting that those who are, can and do are wrong and they (the nonMDs) should lead.

They might be right but it would seem to be pretty much a non-story.

Psychologists understandably have *always* held such positions. Either because that's what *their* training says or because its their economic advantage to believe so (or fundamental bias ... If they didn't feel that way they'd have trained in the other camp.

Perhaps it is retribution for the previously announced Apple intent to 'in-shore' some manufacturing (on top of the already cited government industry competitor ties).

Right, so Microsoft should embargo France.....wait ... that would be a Bad Thing? Perhaps they should carpet bomb Paris with free Windows 8 DVD's ... that would teach the French government to toy with Microsoft and it's minions....