We probably can't observe the multiverse. What we can do is postulate how events might play out if, as is suspected, subatomic particles have some interaction with their counterparts in "nearby" universes. We can model the various possible explanations and use the models to generate testable predictions. Assuming that process ultimately rules out some of the models and favors others, we still won't know that there really are multiple universes, all we'll know is that a model that assumes there are, and assumes they interact in specified ways, is a good explanation for the observed phenomena. "Good" in the sense of being as simple as possible, and no simpler, hard to modify without contradicting observations, etc. Other explanations might work, but just be conceptually weirder and harder for us to think about.

Or maybe we'll eventually find something in the model that demands a multiverse, and for which we can find no other simple, consistent explanations. That will tend to convince us that it really exists. Or maybe the theory will show us a way that we can scale up the interactions to the macro level, where we can observe the multiverse. Or maybe we'll find something that contradicts the multiverse... or maybe none of the above will happen and the whole concept is just a game for very clever people.

Many things are possible, that's why it's fun :-)

Maybe. But this is an area in which we actually can make observations that allow us to refine the math, because it does make testable predictions.

That's great to hear. I didn't realize you were a Microsoftie (though it's on your /. bio). I work for Google, but I'm all for strong, healthy competition.

Neither did Reagan. It's easy to find counterexamples.

The many-worlds hypothesis usually used in explaining the oddities in QM doesn't assume different physics.

Yeah, we're fully aware of the problem. Improvements are in the pipeline. I'm not sure I can disclose what, exactly, although there's obviously no reason to be secretive about this.

By that argument, everything we know about stars, quasars, black holes, and virtually everything else that isn't on our planet and relatively close to the surface is all speculative, too. Nearly everything we know about the stuff not immediately at hand is based on mathematical models, calibrated against "observations" which are often very, very indirect and themselves dependent on many layers of mathematical models derived the same way.

I don't know enough about QM and many worlds theories to know how much really is well-supported, but from what little I've read, the many-worlds hypothesis seems to provide a much better explanation of the spooky action at a distance effects we observe than the alternatives.

Yes, the safest approach is to enable encryption just after you get the device (after using it for a few minutes to accumulate some randomness in the Linux randomness pool, so you get a good key). If you don't, totally wiping it is more or less impossible, though the odds of anything significant surviving either the normal wipe or the encrypt & wipe (which probably won't actually do any more than the wipe) are pretty small.

I'm pretty sure that's not true, because it would make device encryption pretty much useless. A glance at the code certainly appears to show that it encrypts all volumes, but maybe /sdcard somehow gets excluded from the list? I'll ask my colleague, who "owns" disk encryption for Android at Google, tomorrow and post a followup.

I'll also note that none of the devices I have handy (Galaxy Nexus, Nexus 4, Nexus 5, Nexus 7 1st & 2nd gen, Nexus 10, Moto X, Moto RAZR M, Samsung Note 2) even have an /sdcard partition, exactly. They all mount their data partition on /data, and /data is definitely included in device encryption. In fact, it and /cache are the primary targets of device encryption (/system doesn't matter).

But there are also those who don't have a blistering hatred, yet still feel that he broke the law and should be accountable. I find these people to be especially common among those who themselves are or have been under legal and moral obligations to preserve US government secrets and are appalled that Snowden essentially dumped a huge pile of unsifted sensitive data on the Guardian and trusted them to keep it secure and behave responsibly.

These people largely agree with the need to publish some of the data, but find dumping all of it to be criminally irresponsible.

I think there are a lot more people like that than those who have the blistering hatred you mention. FWIW, my own take (as someone who once held a Top Secret clearance) is that Snowden's action was necessary, that it was infeasible for him to properly vet and carefully release the data, that the news agencies have done a good job and been responsible, and that whatever damage it may have done is far more than offset by the good that it has done. So on balance I consider him a hero. But I do know a lot of people whose concern about what he did tips the balance the other way, even though they don't "have blistering hatred".

It also helps to be able to set foot on US soil without being arrested. Not a constitutional requirement, per se, but a fairly important practical one. Otherwise even if you win you have to figure out how to sneak into the country and your own inauguration so you can get sworn in -- and acquire the ability to pardon yourself -- before being body slammed to the ground, thrown into the back of a black Suburban and transported to Gitmo for waterboarding.

Absolutely. So, look at the source: https://android.googlesource.c...

That file contains the code that generates the master key, derives the key encryption key used to protect it (using scrypt), stores the protected master key, and configures dm_crypt with the master key.

Some functions to look at:

- create_encrypted_random_key(), which creates the master key (reading from /dev/urandom).
- encrypt_master_key(), which derives a KEK from your password and uses it to encrypt the master key.
- decrypt_master_key(), which does the reverse.
- create_crypto_blk_dev(), which creates dm_crypt block device.
- cryptfs_setup_volume(), which mounts an encrypted block device.
- cryptfs_enable_inplace(), which encrypts an existing file system.

The device doesn't know you're trying to wipe. It knows that you (a) requested full disk encryption and then later (b) requested a wipe. So it can't optimize (a) away. I suppose it's possible it could just lie and tell you "Yep, I'm encrypting" even though it isn't, but that's the sort of thing that would definitely get noticed by security analysts and gleefully published.