It's interesting that the question implies that Linux is leading the charge in defining new APIs. Everything listed has a FreeBSD equivalent that predates the linux version:

cgroups -> jails
udev -> devfs
fanotify, timerfd, signalfd -> kqueue

Of course, the Linux developers decided to reinvent them all making compatibility impossible. I guess you could argue that the Linux versions offer some extra features over the FreeBSD versions, but from a user and developer perspective, the FreeBSD versions seem more complete and stable (see jails vs cgroups).

The grandparent never said it did. They simply stated that it certainly is illegal in the USA and suggested it may be in Canada as well.

CDNs are designed to minimize latency which also happens to minimize distance (and generally cost as well) to the end user. BT _could_ do something similar by measuring latency of requests and preferring peers that have low latency. CDNs just do it all upfront and manually.

That really only scales up so far. It's actually quite difficult to saturate a 1gbps, let alone a 10gbps or 100gbps, link with a single stream. Multiple streams work around some of the problems and allow the full link to be used.

They are never present in the main search results. They are above or to the right of the results.

Direct message. It's short enough that it should just always be written out.

Except that a century ago, stealing blueprints actually deprived the owner of something tangible. It was actually theft.

Not if we are in the USA, the photos are taken from a public place (the street), and they are only for personal use.

Only if that patch includes new functionality or content and the company is publicly traded.

Actually, the USA has no official language.

Intellectual property from other companies generally has to be stripped from the code base and those algorithms reimplemented in a different way. Yes, technically those other companies could open-source their code, but generally they don't. Sadly, that intellectual property is almost always used to get high performance.

IMHO, Cicero's Pizza in San Jose has probably the best NY-style pizza outside of NY.

Typically for graphics cards, the only data sent over PCIe is texture data, vertex lists, and commands. The bulk of the operations done by the card are running the commands over the vertex lists while bringing in texture data. The commands are almost always a multi-pass or pipeline so each vertex will be used in computations more than once. The result is the pushed to the monitor, not the PCIe. So, yes, in general, a graphics card will have more FLOPs than I/O bandwidth.

Seems like we already have: top speed limiters, safety scissors, plastic butter knives....