There's some truth to that, though a lot of things don't actually require a reboot-- even when they say they do. One of the secrets is that sometimes, asking someone to reboot is just a customer support tactic. For example, if I have 5 things to do in the next hour, and only time to do four of them, I might ask one of them to wait until they have time to save all of their work, reboot the computer, and check to see if they're still having problems. I might not expect that rebooting will fix the problem, but if the client is the sort who will refuse to reboot their computer for 3 hours because they're "too busy" to save their work and close their programs, then I've just bought myself 3 hours to sort out the other 4 cases and research what might be the cause of the 5th case. Besides, even if the reboot itself doesn't fix the problem, maybe getting the user to save/close all of their documents will help, if you know it's someone who generally has a billion windows open at one time.

But that's kind of the thing: Formal education doesn't usually train you to think about things like that. Experience does.

As someone who hires and manages tech support workers (and has done so for a few different companies), I can say that the point being made isn't as trite as all of that. When I look at a resume or interview someone, I don't ultimately pay very much attention to the education. The reason why is that most degrees are virtually useless for the work.

I've known and hired people who have degrees related to computers/engineering, and others who have no degree or have a degree in something completely irrelevant to the job. Regardless of the degree (or lack thereof), I'm more interested in experience. If they have no experience, then they're going to start of doing the simplest grunt work while I train them. I don't care if you're a computer science major. If anything, CompSci majors are worse, because they have a lot of bad habits and misconceptions. You could have a PhD in CompSci, and if you have no experience working help-desk, you're still doing the lowest-level grunt-work until you can prove yourself. Once you prove yourself, I don't care whether you have a career.

On a basic level, fixing computers isn't very tough, but experience of how computers actually work in the real world is often worth more than abstract knowledge of how computers are supposed to work, when they work as theoretically expected. When you get beyond the basic level, the job is more about being organized, communicating, prioritizing, and providing customer service than it is about computers.

Well I think the elephant in the room is that we need a lesser focus on "higher education" and a greater focus on "trade schools". In fact, that's what's happening already, in a half-assed way, when people have the mentality "I just want to get a job/make money!" They're thinking of our colleges and universities as trade schools, and those schools are, to some extent, setting themselves up to be trade schools.

The only real problem that I see with all of this is that we can't make up our minds what we want. Lots of people want to go to schools that will teach them a trade that will make money, but call it a "trade school" and those same people think that it's beneath them, that it's low-class. They don't like learning a broad spectrum of generalized and abstract concepts, but they've been taught that either you go to college, or you should work the cash register at a fast-food restaurant-- there's no middle ground. There are professions like plumbing, which make decent money but people think are for stupid low-class people, and then professions like IT support which are considered more "professional" though it often amounts to similar work-- you're a mr. fix-it working with computers rather than pipes.

It's in coherent.

Meanwhile, colleges are actually more focused on research dollars, sports teams, and frat parties than providing either a "higher education" or a "trade education", all of which confuses these issues even more. I'm of the opinion that these things impede each other, and we need to begin to separate them back out. Young people who have no interest in studying anything and only want to party should go to cities and communities where they can get drunk and messy, instead of coupling that experience with "education". We should have minor league sports teams which have no college association, and let promising young athletes get jobs in those leagues instead of taking sham courses in big universities. We should look at how we fund and handle research and see if so much of it should be taking place in universities. We develop respectable trade schools for young people to learn a trade (or for older people to retrain in a different trade) for instances where people are looking for practical employable skills rather than abstract knowledge.

All of these things are achievable if only we could get our collective heads out of our asses. Unfortunately, I have very little faith in humanity being able to do that sort of thing.

Yeah, well here's a hint: the process I laid out isn't just for planning to change operating systems either.

However, the fact is that Microsoft has generally maintained compatibility pretty well between OS versions. If you have software running on XP, there's a good chance it will run on the latest version of Windows. Most of the apps that won't run are apps that were already old when Windows XP came out, or else applications that are very poorly written.

Yeah, tech savvy people who haven't done IT support often don't quite understand this. There are lots of people-- people of all ages and backgrounds-- who have no understanding whatsoever about how computers work. All they know is, "I move the mouse here and I click this button." They don't understand how it works. They just know, "When I want to process an expense report, I click on this button, then that button, then I type in this product code, and then I hit Enter 5 times." Or it might be that they don't even know what they're doing, but they know, "If I don't click on this icon on my desktop every Friday at 4:30pm, my boss gets angry. I don't know what it does, but he told me that there's something about payroll that doesn't work if I don't click on that and type in my password." If you so much as change the colors of an icon, these people are lost.

Know your audience.

Eh.... it depends. I'd be more open to WINE, though for business purposes it's often not worth it to run non-native apps. Running XP in a VM has its purposes (mostly in running a very isolated application), but as a general rule it's not going to actually solve your problems. It's like, "Congratulations! You've gotten rid of the headache with managing a bunch of Windows XP machines! You've replaced it with managing a bunch of Windows XP virtual machines, which is almost as bad, plus now you have to manage the Linux machines hosting them!"

Depending on the environment, you might also be able to host the one-or-two required apps on a remote desktop server.

Well that's the main reason why I finished up by saying you need to start with a pilot program. Make people actually use it to do their jobs. Involve management in the process, and get management to sign off based on the outcome of the pilot program with the power users. It should cut out a lot of the must-have-but-forgot-to-mention functions if people actually have to use it for their daily work for an extended period of time. There may still be some lingering issue somewhere, but at least it won't be gross negligence on your part if you've missed it.

Otherwise a VM

If you want to know how to start, your first step would be to audit all the software that people use to get their jobs done. Once you have a complete list, ask these questions for each piece of software:

1. Does that software run on Linux?
2. If not, is there a comparable piece of software that would have all the functionality we need?
3. If not, can we live without the missing functionality?

If you get to the end of those questions and the answer is "no", then you should probably cut your losses and accept that you'll have to stick with Windows. If you can answer "yes" to at least one of these questions for every piece of software on your list, then select some users to be in a pilot program. You should find at least a couple semi-influential but fairly patient power users and set up a new test machine for them.

With Firefox and HTML5 you do not need Java and Flash for most content (not for any I consume)

Well to quote the title of this thread, Good for you. That's not the case for a lot of us. Some of us view content or, even worse, have jobs that require we use apps, that require Flash or Java. And regardless of whose fault it is, I've seen content that renders properly in Adobe Reader that does not render properly in other PDF viewers, or PDFs that use forms/signatures that seem to only work in Adobe Reader.

So *yay* for you. You have nothing better to do than secure an outdated OS by limiting what you do on that OS. Some of us don't have time for those kinds of shenanigans.

Yeah, great. It's true, you can secure even the most unpatched and insecure computer by deciding that the user doesn't actually need to use any of the features that could lead to security problems. Take it a step further: I've decided that users don't actually need the Internet, so just unplug the Windows XP machines from the network entirely. It's much more secure that way!

I deal with the same thing, but the reality is that it is very often a result of (A) user behavior; and (B) outdated software. That outdated software isn't always the OS, but often it's Adobe Reader, Adobe Flash Player, or the Java plugin. If I can make sure all the software on your system is completely patched and up-to-date, I will drastically cut your chances of infection. Don't even try to claim that having the latest patches for Flash/Java don't help your security. And how long will Adobe/Oracle support Windows XP once Microsoft drops support?

There's still the issue of "user behavior", which can be mitigated by using certain policies, but can't be stopped without severely inhibiting users. Still, by keeping things up to date, I can cut the chances of infection for all but the most persistently stupid users.

Well in that sense, you could be running 10 different AV packages, running constant scans, and still be running a rootkit. It might be an unknown rootkit that is very good at evading detection.

Short of that kind of paranoia, I'm fine.

If you're a business that is relying on an agilent scope (or whatever hardware) to keep the business operating, then either budget to buy a new agilent scope when support ends, or budget to pay for someone who can offer support.

The problem is, when the hardware (or software) solution will only run on a closed/proprietary system, you don't really have the option to pay for support. If I had the money budgeted, I could pay a couple of programmers to backport updates and patch an old version of Debian that has no official support. Or, if the original system ran on Debian, I could get the programmers to figure out what needs to happen in order to get the system running on a new version of Linux. Even if I have to pay programmers to keep it working properly, keep it up-to-date, and keep it secure, I have that option.

Not so with Windows. I guess you could pay programmers to try to hack patches back onto Windows, but that seems like an awful mess.

So if your point is, "I'm an individual hobbyist who wants to be able to run my little thing and I can't program!" then no, being open source doesn't help you. I'm talking more about businesses who complain, "I built my business around [insert hardware/software solution here], which only works on [insert obsolete closed proprietary platform here]. I can't afford to buy [new hardware/software solution], so I have to keep using [obsolete closed proprietary platform]." Every year I deal with these things, my sympathy for that situation diminishes, especially so when there are FOSS alternatives. Either way, if your business depends on some form of technology, have a strategy and a budget for either maintaining it or replacing it. If you can't budget for that, then you don't have a sustainable business model.

I think this is actually a good argument as to why those people shouldn't be using Windows in the first place. There may be other arguments, e.g. "On a practical level, they had no choice," but I just want to point out that this is exactly the kind of thing FOSS advocates are talking about for years, while most of the world dismissed them as being paranoid.

If you're dependent on specialized hardware, and you will need ongoing support over years or decades, you might want to look for a system that uses an open source driver and/or runs on Linux or a BSD variant. It doesn't immediately solve all of your problems, but it leaves open the possibility that someone somewhere could provide maintenance or an upgrade path. For as long as you're buying specialized solutions that require a specific version of proprietary software, you're at the mercy of the vendor who sells/supports that proprietary software.

When you're planning to buy or build a system, you should always plan ahead. It's amazing to me how many people will buy or build huge expensive systems, build their entire business to depend on those systems, and yet have absolutely no plan or budget to maintain or replace those systems. So many toddlers building empires out of sand.

Agreed.

Thing is, that does sound crazy to me. It sounds compulsive and anal retentive. I wouldn't be surprised if he also only operates his computers while wearing a tinfoil hat inside a Faraday cage that he built in his basement.

You know what I do? I install a modern operating system and pretty much leave it alone. I have never been infected, simply by keeping up to date and not engaging in high-risk behavior. I'd rather spend a few dollars now and then than sit around re-running security checks, but I guess I'm not retired and I don't have the time to be a kooky security nut. I know, someone is going to bash this post because Slashdot has a lot of kooky security nuts, as well as a strong contingent of people who like to hack together weird solutions for what may be non-existent problems. And that's fine as long as you're doing that because you like doing it. Just don't go pushing it as a good idea. You're making everything more difficult for those of us who have to support these things for a living.

The best strategy for most people, especially in terms of doing work for your business, is to stay relatively up-to-date with supported hardware and software.

I don't know why you're being so obtuse about this: You don't have to wait until the population is completely inbred and then suddenly have all these babies from the embryos. You could start selectively introducing new genetic material within the first few generations. My point is that if you're bringing along 20k extra people who will make the trip harder, and you're only bringing them along in order to keep diversity, then why not just bring their sperm and eggs instead. Introduce them into the population as you go if you need to, but I don't think people will become crippled by genetic defects after a couple hundred years if you have a few hundred people.

What, did you write the article? Are you offended that someone might disagree?

Yeah, well in a catastrophic case, you might lose the whole ship in any case, depending on the nature of the catastrophe. Hell, what if you had a catastrophe that meant only 150 people survived the trip? I bet those people would wish they had embryos to introduce genetic diversity.