In my organization, they have found that most of the XP software will work on Windows 8 32 bit. Thet will not help if there is dedicated hardware, but for a lot of cases, it is simply about testing the software on a new OS.

It also depends exactly which software this applies to, Not much help if all of the "trival" is easy to migrate, but the "critical" is difficult/impossible.

However with your definition of $100 USD, cost to upgrade OS from XP to Windows 7, as being "an arm and a leg"

It';s only 100 USD if your time is free :)

Further, Microsoft can support Windows XP, they just want more $$$ to do it (so, if they can do it for one company, and the goods they're selling are infinite, why can't they for all the rest?). If they offered a path to upgrade that didn't cost an arm and a leg, they wouldn't see this kind of lingering on XP that they do.

The current "upgrade" path is one of reinstall from scratch.

If they spent a little more time streamlining their upgrade process and provided proper support for older binaries, maybe. Try to run a Win16 binary on Windows Vista+ and see what happens - hell, even binaries officially supposed to run under Windows 8 won't.

Not just Win16 binaries are affected. Also refusing to run can be less of a problem than running, but doing different things some of the time. Desling with such software can rapidly become expensive.

It costs a lot more than a new PC to upgrade thousands of PCs. Imaging, deployment, backup/restore processes for the end users is just the beginning. Upgrading dozens, hundreds, or thousands of individual customized applications to be compatible with Windows 7 is an absolute nightmare. I know all about this just from upgrading my relatively small workplace from XP to 7. It was a fight just to get core, mission critical apps to work with IE 9; 10 and 11 are out of the question. Lots of cash to vendors and app support folks, lots of cash to deployment specialists, lots of overtime. Adds up to a LOT of money.

XP to 7 (or 8) is really a "migration" than an "upgrade". Depending on a lot of factors it may or may not be the easiest (or cheapest) migration to make.

Much more than that, obviously. You don't replace the operating system, reinstall and develop specialist applications for £5 a PC.

There's dealing with changed behaviours. Especially any which are poorly (if at all) documented.
Subtle differences can easily push up the costs of a migration.

Someone else mentioned DRM for old software that you canâ(TM)t virtualize, like those old printer port dongles that were required to run some software.

There are also cases where vendors throw hissy fits over any mention of running their software in a virtual machine, even if there are no specific hardware issues.
As well as systems which are "embedded" or otherwise tied to specific pieces of hardware.

But the updates will only be created for EN_GB and there'll be no way to translate that into american english.

You mean Microsoft actually have an EN_GB version of WinXP? (And presumably IE)...

I'm afraid its actually *you* who is full of shit in this case, as the Sales of Goods Act 1979 and its amendments are precisely what I am referring to, and as I have intimate knowledge of that act and its various legal successes, I can safely say that you are full of bollocks.

The laws in question only cover "goods and services". Whilst a software licence might have some of the attributes of both it dosn't appear to actually be either, thus falls outside the scope of the legislation. Also most of the protection excludes "business to business".

Sorry, thats a load of bollocks - the NHS has had over half a decade to do something about their situation and they failed,

The NHS isn't a single entity. With various forms of part privatization and outsourcing over, at least, the last 30 years having contributed to this. Such "fragmentation" is likely to make any form of computer system migration more difficult.

One would think that, like filing cabinets, desks, etc., functions like email and document storage would converge into something very durable and unchanging. Instead we've accepted the notion that software vendors should throw it all away with each version update and start over, rather than patch and maintain software so that it converges into something better with each update.

In addition fundemental things can change drastically between different versions.

If, going forward all the plants were of an identical design...wouldn't that make things a bit simpler? Right now it seems the goal is to keep these ancient dinasaur reactors running (which does make short-term economic sense...). But wouldn't a more "monolithic", unified set design standard cut costs and ensure things were safer?

Or it might make things less safe if a flaw was discovered in the design in the future.

Or the new ones are too pessimistic and rely on theoretical possibilities that never can or will come true in reality, but we choose to err on the side of caution.

Assuming they don't also overlook more common/mundane risks. Or, even worst, attempt "risk assessment" by some kind of "box ticking".

Chernobyl dumped 5% of the reactor core materials, one million cancer deaths were predicted, it's been 25 years, why can't the anti nuclear pundits produce a scientific peer review study showing at least one hundred thousand actual deaths ?????

The process of "peer review" dosn't in itself make a study correct. Especially where this issue is mixed in with politics. When tends to be the case with "X causes Y premature deaths". All too easy for any study to be an attempt at "proof"... Best to be able to see the bodies or at least the death certificates.

As for Fukushima. Fukushima is the story of a freak Tsunami that was mutated by the anti-nuke community into a "nuclear failure".

Assuming that the "anti-nuke community" didn't contribute to the problem by making it more difficult to replace the old reactors an/or move spent fuel off site.

how do you deal with the inevitable "Hey, you're in my seat" dilemma?

Not all airlines assign specific seats to specific passengers. Some even charge for specific seats.
A more obvious question would be if the crews do a "head count" or not.
Wonder if the article the article should have said "Schengen Area" rather than EU.