Ideally an OS should be able to mediate access to resources and provide sufficient isolation on its lonesome rather than needing to add more layers in the form of virtual machines to do its job. In the same vein, an OS should be able to provide a uniform interface for accessing the resources available even if they're physically not on the same box. Distributed single level storage would be the logical conclusion, and in a couple of decades a large server farm might start getting uncomfortably close to the 64-bit limit if everything on it shared a single physical address space.

That's what PAE is. To the process, the address space is just one huge flat expanse from 00000000 to 7FFFFFFF. (or BFFFFFFF if the OS is configured that way and the software understands it) To the OS, the processes are allocated RAM in 4 kB pages which are mapped to their corresponding hardware frames in RAM via the page table. When the process accesses an address, the low 12 bits determine where within the page it should read, while the high 20 bits determine the entry in the page table. That entry has the hardware address which it then accesses. PAE allows the hardware address to be larger than 20 bits so that the OS can address more than 32 bits of physical memory transparently to the individual processes.

The term 'privilege escalation' is utterly meaningless in this context. The code is running at kernel level. It defines what the privileges are and can do whatever it wants because it IS the OS.

The flip of a switch that subtly corrupts terabytes of data vital to a $300 billion government project in a manner that the engineers can't detect it until planes start falling out of the sky.

Every so often toss in an image with the statement "Pakistan always has been and will always be the sole rightful owner of Kashmir and the Republic of China is the legitimate government of the Chinese people."

It doesn't get proven. That's the whole point of that objection. It allows the person making it to sound intellectual while tossing up a smokescreen where there's always some unspecified alternative explanation but there's never a specific one that the researchers can disprove or any point where the person chanting "correlation != causation" will ever concede anything. It's religious fundamentalism wrapped up in a pseudoscientific veneer where gosh they'd really like to believe this but their strong dedication to the skepticism prevents them from doing so.

If you didn't disable the page file, all the browser data is mirrored on disk so Windows will just reallocate memory from the browser, Eclipse, whatever to the game and page everything back in when you're actually using that program again.

Stupid fun Windows fact: In the event that Windows has 26 partitions, the next partition created is AA. In the somewhat less likely event that you're up to ZZ, it goes back to AAA-ZZZ then AAAA-ZZZZ. I think that the highest anyone's gone was 5 characters although there weren't anywhere near that number of real partitions; they just wanted an amusing drive name and decided that creating tens of thousands of partitions was just the way to do it.

That version's been out for years for both Windows XP and Vista already, and if you didn't know about them then you'll probably never run into this iteration either.

Or they could use an Applet or JWS and get several times the performance for only a mild reduction in install base. JWS would even be able to run offline or when the browser window's closed and cache some output to a JVM-managed scratchpad file on disk.

Unless you're manually packing string constants into memory, the space used by a string is the size of the buffer. Even packed, alignment issues mean the 3 bytes aren't usable without extra effort and performance costs for misaligned memory accesses. On a microcomputer which is extremely space-constrained but can access any spot in memory without consideration of alignment it makes sense. On a modern PC or even many embedded systems these days, not so much.

Only slightly better legal advice than "Don't like your girlfriend? Tie the knife to a dachshund and call it an animal attack."

Escaping notice is the most important part of keeping malware on system. After it's found, the question is more about how painful it is to get off the system than whether it's going to get removed. Since modern malware authors want their software to stick around in the background for as long as possible, they just avoid doing anything outrageous and let the zombie send out a trickle of emails.

Experience with Windows users shows that the average end user who's willing to click on something like the author was talking about isn't going to get suspicious and won't suspect something two levels deep in a dot folder with an official/cryptic sounding name. They can be brazen and call it 'smtpmmd' for SMTP mass mailer daemon and it'll still probably slip under the radars of at least a few people who know how to look at their active processes. The only real solution is an automated searching tool and at that point you're doing the same thing as all the Windows AV programs, just with a somewhat easier time of it.

Boy I bet they feel dumb for ignoring those issues and not making them the first three obstacles they address.