I was concerned about this as well, and frequent crashes on my laptop (battery empty) can ruin a file system (I have made some bad experience with reiser4 in that regard).

However, I tried it, including forced poweroffs while writing, many crashes, etc., and it is fine. You mount the encryption, and recover the file system as usual, and the encryption layer does not influence the recovery at all.

I can recommend ext4 with LUKS (cryptsetup). It is very easy to set up for a single partition. You can choose AES or TwoFish (512 bit key).

The other thing I was worried about was read/write throughput. There is a benchmark utility that will tell you how how different cyphers perform. However I have never noticed any difference when working with encryption, probably because data comes in blocks and is cached efficiently by the kernel. Today, I do not see any obstacles for encrypting some partitions.

I've got a couple of Macs along with the usual iPhone and iPad combination and while the upgrade to Yosemite had some issues for me these were entirely related to older software with incompatible kexts and resulting kernel panics on a regular basis. Once I found the offending item the panics stopped and the machine affected has been fine since. I was a beta tester on Yosemite and have used it as my primary OS since the start of the beta and all in all it has been way better than the likes of Tiger which was terrible when it came out (you could only open a shell once out of about five tries up until 10.4.7 which fixed it) so I don't think the quality is getting any lower in my experience.

I've used the Mac's built in Disk Utility multiple times to image my HDs onto new SSDs and then they've been a simple drop in replacement but you're right, the activated commercial software may well refuse to run.

Adobe uses FLEXnet I believe and that does take a fingerprint of the hardware specifically to stop users from duping discs. The company I work for also uses FLEXnet and when a customer encounters this, their software won't work on the new drive and they have to reactivate on the internet with their original license key and often they run out of activations in which case they need to call our support line so we can identify the old dead activation from their old drive and release it for them.

No it's not because you can add your own, or third-party repos, without needing authorisation.

To install python libraries like scipy, matplotlib, etc. Apparently that is such a pain in MacOS, and there are so many half-assed distributions methods that you can really botch your system. It makes sense to run a Virtual Machine with Linux on it.

So you have not read any, and have no idea what you are talking about. Start with the open source definition (opensource.org) and the Free Software Foundation (gnu.org).
https://www.gnu.org/philosophy...
http://opensource.org/faq#free...

You are making, unintentionally, an excellent point that one should refer to gratis software and libre software.
https://en.wikipedia.org/wiki/...
GCompris is always libre software, but sometimes not gratis. That is OK with both the FSF/GNU and OSI.

The president on Friday argued there must be a technical way to keep information private, but ensure that police and spies can listen in when a court approves.

If the court approves, they can just go and obtain the computers. That is already solved.

If the hard disk is encrypted (very rare I suspect), the expectation of legal costs or indefinite holding at Gitmo without any trial are already there as motivation to comply.

No, better spying is not what we need. It destroys our freedom of speech and quality of life. We need due process. We need protection of all those not proven guilty yet, because it could be any one of us.

Just run the cable through a guitar amp. I mean, what can possibly go wrong?

People have been designing virtual networks for decades. I2P is well advertised on Freenet, itself a well-known secure network.

Nothing new here. The security and reliability of none of this software is proven, it may not even be provable due to the distributed nature. That reduces the problem to one of how many people you're ok with knowing what you're doing.

They probably try to avoid torturing with Tasers. This happens when the Taser is activated multiple times or for extensive durations (e.g. 3 minutes, causing death)

That would be great, but it is currently not possible to run a mobile recording camera 24/7 with the batteries available today.

I honestly doubt any severe repercussions will occur, the DMCA is too weasel worded. Defamation is another matter. Accusing a company like Atlassian of hosting pirated porn is a serious commercial matter. (Slandering open source developers is another matter, freedom of speech and all that, America hasn't really grasped the concept of reasonableness and balance.)

Accusations that are clearly defamatory against a commercial entity can harm political donations, jobs in battleground states, and inflict restraint of trade, on the long run-up to a major and likely to be bitter election... That is clearly not going to fly with elected judges and elected political representatives.

The question is whether legitimate businesses involved in legitimate trade will simply ignore the action or file for defamation. Winning or losing doesn't matter, most of the porn companies are probably small enough that bad publicity and legal fees will cripple them. Obviously winning (even if by default) would be better, it would create case law on the issue.

The problem with DMCA is that we've been here before many times. And there have been DMCA cases the industry has lost. Yet nothing has changed, no precedents have been set, no behaviour on the part of industry or takedown farms has been modified. You'll have to do something new.

On the deirious mode a really well developped being would eats the whole universe and and in passing eats all gods. Consequence: it would enter an infinitie loop of eating itself, no need to look for food again.

From BG blog one can conclude that the author belongs to the category of people unclear about the difference between a quantity of energy and a rate of energy production. To his excuse the common poor choice of kWh instead of the SI J (Joule J, 1 kWh = 3.6 MJ) as energy unit is just making energy discussions more confusing.

Noise could also be proportional to the unresolved intensity. However, you can see that the dots are actually round, and thus resolved stars, and not simple individual pixel noise.