Here's an interesting idea: given that they do have a monopoly over that content, how about we start enforcing antitrust/competition law on it? That would mean no more requiring a foxtel (cable, for non-Aussies) subscription to view a single piece of content. You could achieve this with the existing CCA, which defines markets (for the purpose of defining a monopoly) as having some element of substitution between products within them, though it would definitely be a novel interpretation.

Alternatively, we could recognize that exclusivity agreements are anti -competitive, and therefore illegal under CCA s45, but that's going to be much more disruptive.

You have a vulnerability in your config - no unprivileged process should be able create a DoS attack so easily. I suggest you take a look at /etc/security/limits.conf.

(Ironically, I set a limit on my system due to a similar bug in Kontact, which has since been fixed.)

For Windows, that's quite plausible. For Linux - not a chance. X and journalctl might crash, but the rest will generally keep going, as long as you don't have have a COW file system. It's actually not that different to how the system handles running out of free space (a much more common occurrence, particularly if you don't keep /var/log on a separate partition).

You can get most of those with something like the Moto X (easily rooted, all bands, microUSB). Add the right case for Qi + strap loop.

Newsflash: *most* USB devices lack serial numbers.

Of the 18 USB devices plugged into my PC, only 2 of them have serial numbers, and they're the same model. The USB hubs built into the motherboard also have serial numbers, but those can be identified by PCI address so that's not as useful.

I have a Linux multiseat config, and the only way to reliably recognize seats is to make sure you always plug the hubs into the same ports.

The problem arises when the internet is wrong. For instance, just yesterday I was setting up Postgres* (something I haven't done before) and ran into an error about peer authentication when trying to run psql as root. If you google this, pretty much all the results tell you to set the authentication type to 'trust' for the postgres account, which effectively disables any kind of checking for local access to an account with superuser permissions to the database, creating a privileged escalation vulnerability. It took me about half an hour to find out about user name mapping and get it to work.

And as someone pointed out the other day, this is especially rampant with certain languages. e.g. PHP.

Googling a problem is certainly an important step for getting ideas, but sometimes it's worthwhile reading the docs and figuring it out from first principles, because quite often understanding and checking the code can take as long as writing it.

What's wrong with gitk?

Voat seems to be doing pretty well...

Counter-point: I just spent $500 on a graphics card, and my gaming system is single-boot Linux. Now to be fair, I do play quite a few games through WINE (though fewer than I used to), but the proportion of games which are Windows only and unplayable in WINE isn't as high as it used to be. I did get an Nvidia card (GTX 970) though, mainly because AMD's drivers have such a poor reputation under Linux.

There's plenty of hardware that doesn't work well under Windows too, either because of driver bugs or because the hardware itself has a design flaw (e.g. the GTX 970 memory architecture). Ultimately you need to do some research no matter what your OS is, if you're going to build the system from scratch. (And if not, the Steam machines are equivalent to a pre-built Windows system.)

I'd argue it still is, to a large extent. I'm fresh out of uni, but I (and several of those I graduated with) used Vim/Emacs as their (DIY) IDE of choice.

(FWIW, my course spanned Java, Python and C, and only for Java did we use an IDE (Eclipse).)

I run Sabayon, you insensitive clod! :P

This is true of mostly just C, though it does get used a little in C++ as well. (To be fair, C-family languages are the majority of popular languages.)
Most high-level languages use exceptions, and Haskell (and to a lesser extent, Rust) use the Maybe type so that the type system forces error checking to occur.

Actually, there'll be brand new Linux drivers for this card in Linux 4.2, which confine the binary blob to user-space. Whether or not they qualify as buggy remains to be seen though...

Judging by the entries in my log files, a very simple but effective mitigation technique would be to run wordpress under a path like /blog instead of the root directory - most attempts are just blindly searching for vulnerabilities at the default path.

Honestly, even with systemd[1] Gentoo is actually a pretty awesome distro for a server; definitely better than Debian. The compile-times are less of an issue since most of the really large programs aren't ones you'd install on a server (chrome, libreoffice), and being able to install any of the last N stable versions (sometimes side-by-side using slots) of a package is pretty convenient when you have third party software. The way configuration files are updated after an upgrade instead of during is also much more convenient, and the stable branch is also significantly more up-to-date than Debian's.

[1] While I use systemd on my own system and find it works well for me, I can easily understand why some people would prefer to stick with one of the alternatives (at least until all the bugs get ironed out).