He's looking for an excuse after the fact since the time expenditure would be less than he spent on the posts complaining about the time expenditure.

A bit less tinfoil, how about there being far less care taken with the old version than the new?
I'd add hardware requirements creep to that where the testers (don't laugh, I'm sure some testing is going on) have far more capable hardware than is typical for the install base. Not slow with an i7, SSD and 16GB or more of memory could be unacceptably slow on more typical Win7 hardware. Two cores and 2GB just don't seem to do it any more when it certainly used to for office tasks.

Does your other OS hold on to outdated versions of system files for compatibility reasons, like windows 7+ does?

Of course it does.
There's a thing called version numbers of libraries that predates MS but they didn't decide to go that way until relatively recently. It's a way of being able to retain old versions of system files for compatibility reasons without slowing down the system searching for stuff. That is why linux, solaris, oracle, mac, *bsd and everything else apart from MS can run old stuff without a great deal of mucking about.

New files get added during the day and old ones get modified so if you want to protect a malware-prone environment effectively there is going to be a lot of scanning activity during working hours.
An after hours scan reporting that a lot of stuff got trashed by cryptolocker at 10am is of very limited usefulness.

Or don't turn it off EVERY SINGLE FUCKING DAY, complain about it being slow to start, repeatedly, perhaps three times a week or more, but pointedly ignore advice not to turn it off EVERY SINGLE FUCKING DAY. I've got a few people who do that who don't have the patience of the above poster to have it turned on some time before they need to use it. Sometimes I think it's a work avoidance tactic to justify hanging around and chatting in the mornings.

Bucketloads of memory reduce the pain of running a malware-prone system since those on-access scans can be on a cached copy of the file instead of having to hit the actual disk twice.
Now we have desktop motherboards that can take 32GB of RAM relatively cheaply so what would have seemed excessive in the past is now a viable workaround.

Good point - the failure mode of the VLC media player (and many others) when they can't keep up is to keep the audio running and give up on rendering frames.

A window trying to draw should be the highest possible priority on desktop workstation and should not be treated like other processes. There is good reason to keep other processes waiting while I/O completes but window drawing operations should never be kept waiting

One of the more odd design choices in MS Windows is a good demonstration of where you do not want the rendering to have priority. Click on "control panel" and it starts drawing a lot of little icons, then it finds a lot more things to add to the list and redraws with a lot more icons. If a user attempts to click on one of the icons in the several seconds (yes that long - how fucked is that?) when it's rendering stuff and reordering it then they are very likely to click on a icon that was not there when they decided to click and end up opening something different. A sane way to do it, as done in many other parts of the MS GUI, is to make a list AND THEN present it to the user instead of a slooooow interative process. The desktop starting up is another example like that where you can see controls but can't use them for up to tens of seconds depending on how much stuff is loading. A sane way, which as far as I know is used in every non-MS computing environment, is to have some sort of splash screen or indication that the environment is not ready, then it provides the controls at the point where you can actually use them. A marketing choice to have X seconds to the desktop and cheat by drawing controls before the user can actually use them means we have an interface that frustrates and confuses users and the public perception of the relaibility of computers has been going downhill over time.
So IMHO rendering should wait until the user can actually interact with the rendered thing. Putting it there early is frustrating for the user. Nobody wants to click six times on a thing before it's ready then eventually get six instances of it when all you wanted is one as soon as possible.

Linux distributions assume you really want to give priority to server applications

Since the year 2000 Microsoft has been doing that too because it turns out to be a really good idea for anything more multi-purpose than a game console.

You start off with a machine that runs like a greased greyhound and once the users put flash, dropbox and other non-work related shit on there it slows to a crawl. Meanwhile identical systems with linux or similar on them and only work related stuff stay at around the same speed unless disks start filing up or new software needs a bit more grunt than the old machine can provide.

Even as an experienced systems integration engineer, I would need a few hours at least to develop a plan on how to do that

Joking, newbie, selling yourself way short or completely and utterly fucking useless - what is it to be? The amusing bit is the condescending crap on the end about home backups when the situation is that if you are responsible for the gear then you are failing in your duty if you cannot do a bare metal restore of critical systems AND talk somebody with minimal experience through it. I've been there and had a complex pile of stuff only I knew how to restore properly, but I did my job and got it all down to a procedure just a few lines long with simple steps and it gets packed in with the tapes.

you have to know the overall engineering plan to set them up.

So you refine the plan so that a monkey can step through it and you document it well enough that you can read it to somebody two minutes after waking or someone with limited experience can read it themselves.
FFS - it's far easier now than it ever was before since we can boot stuff off USB drives and then remotely populate their disks with what was on their before instead of reinstalling before restore.

Since you laid on the condescending crap it's time for me to ask you a question. As a "systems integration engineer" shouldn't you be considering things from an engineering approach of improving the system that is broken instead of an ad-hoc basketweaving approach of the technician just doing what seems sort of OK a different way each time while waiting for someone to write procedures for them to follow? I don't consider myself an engineer anymore since I've been exclusively on the IT side since 2000 but I do still apply the approach that I used on engineering problems, something you self-declared engineers who do not have their title accepted by a professional body should consider if you want to be taken seriously.

The seawater cooled plant would have had brass tubing that is very resistant to salt water corrosion, but since that brass is worth quite a bit it would have been removed and sold when the place was shut down. Of course if you don't have to worry much about heat there are plenty of types of cheap platic tubing that can do the job.
The only "irony" here is that power stations typically don't have a lot of floor space for the amount of area they occupy once you pull the boilers out, so maybe a one floor datacenter taking up the same volume as a six floor one.

Yet you went on with shit such as suggesting your guesses were far more valid than me getting information in person from someone who had actually been there at the time. Why do such a thing when you should know better than to do so?
Don't just boast about your background use it.

Sometimes the recruiters just aren't good at finding the right people.

Since they often don't know what the key words actually mean it's very difficult for them to do so. That's the problem with using general recruiters instead of getting someone with a skillset related to that of the person you want to employ involved early in the piece. If the perfect employee lists a skill of which the desired skill is a subset the general recruiter is going to reject them because they only know the key words and not what they actually mean. If your degree says engineer and the recruiter is looking for a programmer many recruiters are going to reject the application without even getting as far as considering experience in previous employment.

They are all bullshit and easy to game so it's just an indication that the HR folk have never been anywhere near a psychology textbook or even listened to someone who has.
It's also solving a non-problem. Having a mixture of personalities in a workplace is better than having a monoculture - for instance Enron with hard driven floor trader types that had never grown up was like a high powered fighter jet pointed directly at the side of a mountain. Selecting for such types resulted in an epic failure with a bunch of idiots who didn't understand what mistakes they had made even after the place imploded spectacularly.