Liquids aren't hard to defend against, all it has are two wires with +5 volts and ground and a resistor across another two lines to tell the device its a dumb charging terminal (that doesn't need to be asked to draw more power). Some simple epoxy around the wires and a current limiter (common in just about every usb setup anyway) will take care of conductive liquids (by preventing them from doing damage to the electronics with the limit until the liquid drains out of the port).
Chewing gum stuffed into the USB port is likely the most common and hardest to solve problem there. I presume they would have designed the contacts to be resistant to chemicals that dissolve gum to allow cleaning, but still not going to be pretty.
Other then that, its a big metal box, assuming the solar panel is covered in suitably tough plexiglass I don't see too much in the way of likely damage (but I'm not a vandal, who likely have more experience in how to cause problems that aren't easy to fix).

It looks suspiciously like the smaller armrests placed in the middle of benches in order to prevent the homeless from being able to sleep on the benches, rather then just a poor design decision; but I don't really know what the thought process behind the decision was. I agree with you, a poor design that is also ugly.

They are talking about the GPS ground stations that monitors the GPS signals (and is programed with its exact position and altitude) and determine what corrections, if any, need to be made to the GPS signals (so that what it knows to be its correct position is the same as what its GPS receiver is telling it)
Russia wants similar ground stations set up in the US for their GLONASS system, which I think is fair (and good for users of navigation systems, if not for the US military which would like to be able to turn off Russia's navigation systems).

Where there is an IT team to provide support using SSL client certificates will prevent (and detect via server SSL logs and client errors) fake certificates.
When enabled the client will sign (using their client cert, generally with a site-specific internally managed CA) all the communications after the key negotiation finishes, so if there is a middle-man that modified the certificate/keys the server will see the clients signature of the communications as incorrect (as the client and server wouldn't agree on what the communications were) even if the user overrides the SSL certificate warning or an attacker (or employer, or user, or vender) adds a fake/compromised CA to the trust store.
Doesn't work for sites without a support team to work with users and investigate failures or in cases where the internal CA is compromised, but for the highest of security needs its more effective then using Flash.

Reading the wiki on both desalinization via boiling and reverse osmosis indicates the vast majority of the energy in both methods (ether boiling the water under a vacuum or pumping at a high pressure) seem to be independent of how much salt it has to remove; so reducing the salt content by a few percent won't reduce the energy consumption (but will flood most existing plants and require more energy to have them rebuilt).

For my Prius there is a metal ring around the start button (and a mechanical key built into the key fob to open the doors) and if I press the fob on the ring even without a battery in the key the car can power it enough to authenticate the key and will then allow me to start the car.
The price of the keys, however, is indeed unacceptable.

My (limited) understanding of how drugs like the ones talked about here work is that they increase the number of "ticks" that the brain records so it thinks more time has passed, rather then actually speeding the brain up.
So, if the intent of work is to torture people with work then it might be effective, though afaik said people wouldn't likely get any more accomplished per earth-year then anyone else. Perhaps they would waste less time thinking that more wasted time has passed being wasted or something of that nature depending on what it covers.

I don't work from home all that productively, and do much better from a workplace that is split from my living space (and don't want to pay for the space to have an office in my apartment).
So for my uses it would take more innovation then virtual conferencing tools.

Currently they just don't fix problems in cars software unless there is a recall.
There haven't been any patches for the security holes associated with the electrical impulses causing doors to unlock (a patch requiring the door controller to get a cryptographic hello should do the trick), nor the issue allowing one remotely take control of a car, never mind the assorted annoyances that a software patch could fix.
If they were actually able to remotely patch a car there would be more questions about why they aren't making the patches, and they would rather not the focus be on them being cheap.

One problem with ICANN now is that they hold the root DNSSEC keys, so anyone who controls the strings of ICANN can spoof otherwise secured DNS records (and the associated SSH/PGP/HTTPS key pinning done with said records). The NSA, for example, I'm sure would be interested in the ability to man in the middle domains that are seen as important (ones that someone bothered to sign with DNSSEC).
I'm also sure that the GCHQ is equally interested in getting their hands on said keys.

I agree, however the current uses of flash are:
1) Videos
2) Copying text to the clipboard
3) Cow-clicking Games
The first could be moved to HTML5 but DRM (or in YouTube's case advertisement functionality) are slowing that down. DRM by definition can't be open source, so we ether have Flash or HTML5 DRM extensions (that are likely to be almost as bad, in addition to not being maintained and having security researchers yelled at as pirates instead of fixing the vulnerabilities).
In the second case some browser extension could be developed to allow copying only on button click like the popular flash applet does for that, better then having a full programing language for it.
The third case could make use of a foss solution, but with the first two having better options and devs already set in the ways of Flash its unlikely that there would be enough of a market for that to become substantial.

Rather then attaching the public key, a system such as GPG's pka that publishes keys for e-mail addresses in DNS via DNSSEC signed records is likely a safer alternative against modified keys. It also allows the first e-mail between two people to be encrypted (as the key can be found via a DNS request).
PKA works now, but the clients have to be told to use pka manually, so its of limited value in its current state until adoption gets a little wider. Sadly leaves GMail and friends out in the cold (unless they offer a key publishing service to their DNS), but works well for privately controlled domains (since commercial webmail can't really be secure anyway, its as good as I think we are likely to get).

My main issue is sites that SAY they allow anything, but don't. There are a lot of sites I've run across using keepass that don't have data validation cheks when changing the password, but when the password has some symbols in it, can no longer be logged into. I've had one site with the combination of this and a password-retrevial instead of a password reset feature, thus breaking the account completely.

We could have people working at the polling stations who act as proxies to assist voters. The voter tells the worker who to vote for, then the worker places the vote.
Because we are short on money, they canidates should pay these workers, and decide how many and where they work.
I'm sure self-regulation will work fine for this, so faud won't be an issue.