Here in Norway, the maximum sentence is 21 years, and doesn't stack beyond 21 years.

First-degree murder has a maximum of 21 years. So, you could have two guys in jail, both serving 21 years, one which murdered one person, the other which murdered say 69.

Now, I believe that taking 69 lives in cold blood is significantly worse than "just" one. However the sentence does not reflect this.

So while 400 years for one murder is a bit much, sentencing the second guy to 69 * 21 = 1449 years in prison would at least more accurately reflect the crime he committed.

How many of those years he must serve could be orthogonal, if society wants it that way.

How old was the recording you used? The song/tune itself can be public domain due to age, but the performance/recording will still be protected by copyright if it was made recently.

I simply never assumed my Dropbox files were private to begin with.

While I don't share everything in my public folders, I don't put anything in Dropbox that I don't mind the whole world to see.

Because he put the car in neutral?

Oh wait...

Why can't they use a PRNG to dictate the detector settings? Pick a high quality PRNG, seed it with the first prime number. Run the experiment N times. Restart, seeding it with the second prime number, and run the experiment for N times again. Repeat M times until satisfied.

Then re-run the above with a different type of high quality PRNG.

Am I missing some big clue?

Tried running it in "safe mode" without addon's and see how that goes?

Firefox still crashes for me when it runs out of memory due to buggy javascript in either an addon or on a page. For example we use FinalBuilder at work, and the build control page has a massive memory leak in the javascript (sucky dom handling in web 2.0 crap) causing FF to run out of memory if I leave the page open over night.

Other than that it's been very stable on all the machines I've used it on for many years now (and that's both Windows and Linux).

The faulty code is of course fully deterministic. The resulting output may be completely arbitrary.

If your code contains a state machine and the code has a bug where it in some conditions references invalid or uninitialized memory to determine it's next state, then the bug may appear quite random. In fact, it may appear to be several different bugs.

This is particularly fun when doing cross platform programming, where some platforms always zero initializes memory upon allocation causing programmers who develop on those platforms to think everything is peachy, only to have the code promptly crash or worse when run on platforms which does not zero initialize memory.

Web scale or not, I have no idea. But I do think they've got the best bug report ever in their tracker.

Because we use portable libraries and code, and we have plenty of testers.

To be fair, our GUI demands are very modest. Project is LuxRender, a physically-based open-source renderer. The GUI is mostly just for visualization of the resulting render and tweaking tonemapping and other post-process parameters.

We didn't have a dedicated developer on OSX, more like an enthusiastic user who could compile it but not so much code. Sometimes he could Google some workarounds other times we did a workaround in the blind.

In either case we did not feel comfortable hacking the wxWidgets internals, as we had no OSX experience or ability to debug.

With Qt this isn't an issue, as we mostly don't have to worry about stuff not working on OSX if it works on Windows and Linux.

Yeah, I get that. But given than the OSX offering from wxWidgets was pretty much broken on a constant basis, not-quite-native-but-functional Qt widgets won the day over broken-every-other-week wxWidgets.

I exaggerate slightly, but the lack of proper OSX support was the main driver to Qt for us.

Ironically this is the reason we moved our cross-platform OSS app away from wxWidgets to Qt. The native widgets just didn't work properly and it was a pain to fix. We made the move some 4 years ago or so, and I can't say we've noticed we're missing something...

Indeed! You have two options. Either via network or physical disk. If you do it via network, Crashplan will perform the backup when both PC's are online. If you need to restore you can copy the backup repository from your neighbor's computer onto a physical disk and restore from that at home.

Alternatively you can simply use a couple of USB disks, set up as two separate destinations for the backup set in Crashplan. Keep one at home and one at your neighbor. Once a week or whatever you swap them. Crashplan will automatically detect the disk when you plug it in and start syncing the backup.

The best part of this is that the data is encrypted in either case, and IIRC you can do all this using a free account. The paid options only matters if you also want to store the data in "the cloud".