if someone else suggests it, that's fine. you don't go around asking for people to literally 'idolize' you.

It's a bad sign when these types of reports no longer invoke any sort of shock. It's a part of "Americana" now.

the main thing is that the virtual memory statistics are relatively clean. it seems only memmapping, shared libraries and graphics buffers for compositing produce obvious changes in virtual memory statistics, with compositing taking up almost all of the obvious changes. introducing randomized noise into the virtual memory usage at the kernel level would drastically cut down on this attach vector. it would tie the security of the kernel's randomization to the virtual memory usage.

yeah that is private memory. the article is about an attack vector based on changes in shared memory usage ("signatures") for when the OS does its UI compositing. shared memory is used to avoid copying graphical data when a client-based window manager does its compositing.

you still can't read the shared data of another process but the "signatures" can provide you with some idea of what the top-most activity is on the screen, something you cannot know otherwise unless you are the top-most activity or the system-level ActivityManager. if you know the top-most activity, you may have a better idea of what type of attack to launch.

Old Argument (Younger Dryas): We found iridium (rare on earth but abundant in meteorites), nanodiamonds and magnetic particles covering ancient tools and mammoth remains at sites which we believe are around 12,000 years old. Therefore, we believe a cosmic collision caused the 1,300 year deep-freeze.

New Argument: We performed radiocarbon dating on tools found at the 29 sites described in the Old Argument and found that only 3 of the 29 sites were around 12,000 years old. The tools at other sites were much older or younger. Therefore, the deep-freeze was probably not caused by a cosmic collision.

As a 30 year old engineer architecting and developing 3d graphics engines, I also find these kinds of interview questions worthless and stupid.

so they can purposely botch it and justify the need to have human operators. in case you don't know, BART is currently going through a tense union battle resulting in a few worker strikes and contract disputes.

Nokia has been effectively castrated. It's only a matter of time now for the rest of the company to devalue even more. That means purchasing patents for even cheaper in the future.

Steven Elop works for microsoft. Steven Elop goes to Nokia. Steven Elop restructures and retools Nokia to be a Microsoft shop. Steven Elop cuts Nokia's market cap in half. Microsoft buys Nokia. Steven Elop becomes CEO of Microsoft in a few years (after Ballmer's successor resigns after 2 years). You guys connect the dots yet? I'm sure Nokia has a lot of patents Microsoft wants.

this is a triumph; huge success. (rip clam)

Additionally, any violations of these laws will result in life imprisonment for the board of directors and all executives.

"Dixon crossed the line between free speech protected under the First Amendment and criminal conduct when he told some clients to conceal what he taught them while undergoing government polygraphs."
Assuming the summary is accurate, he crossed the line when he instructed them to lie. There is a big difference between teaching people how to circumvent a polygraph test and overtly instructing them to lie.

quacker.

a direction vector and a rolling average should fix the problem you mentioned about passing over a slower zone.

just throw more big data at it.