Ever tried receiving packages when you live in a smaller apartment building with no reception area or anything like that? USPS generally has access to the actual mailboxes, but USPS and friends will force you to either be there. Your options are: sit out front of the building for hours (I love it when my package is "out for delivery" at 6 AM and gets delivered at 8 PM), go pick it up at a depot (which is not the same as a UPS store, no idea why they can't do that), or just sign the note they leave. That last option isn't available if the sender requests an in-person signature.

So I order something from Amazon, and they required an in-person sig, which I didn't even know until they left the note. After calling UPS to try to just get them to dump the thing out front, they directed me to their website. Of course, their website didn't work because it thought my address was invalid, and their phone support couldn't help me with that. I also called Amazon and asked them why the hell they required an in-person signature to begin with, and their support was equally unhelpful.

The end result? I had to drive for a total of an hour to pick up my $7 cable 2 days later than I should have had it. I'm pretty sure it would have been a better use of time to just go to a store to buy it. Why Amazon required in-person signing for a $7 package to begin with is beyond me.

And then someone tests positive. So what? Just because someone does drugs does't mean it affects their ability to work. If drugs are clearly affecting someone's performance, then it should be pretty obvious without a drug test. A drug test is just a feel-good measure.

That's not just a trend with automation, it's a trend with computing in general. More and more often, the computer thinks it knows what the user wants more than the user. This might be true for an average user, but for a power user it just gets in the way.

The problem I have is that it will probably end up far more bloated than it should be. The less bloated stuff there is, the less of a need for updates there is. Why would an IoT device be vulnerable to shellshock when it should have had a lightweight shell like ash instead of bash to begin with? Why would it have systemd instead of an embedded-centric init system like procd? Automatic updates are actually terrible from a usability standpoint because something can quite literally break overnight.

Drive space is cheap, and so are current backup services. I can't imagine that this service would be significantly cheaper than the existing services, so what's the point? There's unlimited (yeah yeah, "Unlimited") backup services for $5/month, so it's not like there's even much money to be saved.

It's no different than doing the exact same thing over Firewire, but it's a lot easier to hide an exploit in plain sight. When you exploit over something like Firewire or Thunderbolt, it could be a simple "Hey, can I charge my iPhone?". I remember an old exploit that you could do using one of the ancient Firewire iPods. That's a lot different than "Hey, can I plug this random card into your computer?" when you want to do it over CardBus or ExpressCard.

On a desktop, I don't think it would be a problem. If you had a rather standard encryption scheme where you enter your passphrase on boot, it wouldn't be exploitable because someone would have to shut down the machine, stick a PCIe card in, and then boot again, thus losing the encryption key until it is entered again. It's just that laptops tend to have to have more exploitable interfaces that support hotplugging (like ExpressCard and Thunderbolt) whereas a desktop at most might have Firewire.

I'm surprised nobody has engineered a DMA exploit over SATA, considering it's hotpluggable and rather ubiquitous.

I don't know if the newer ones are any better, but my W510 definitely has overheating issues. They basically took the exact same cooling system used in the T510 and tried to use it to cool a quadcore and workstation video chip. The CPU could hit the point where it would start to throttle. Even after putting on better thermal paste, it still gets hot enough to cause the GPU to throttle. Basically any high performance laptop that doesn't either have a huge cooling system or a separate system for the CPU and GPU can run into these issues.

From what I understand, thunderbolt is essentially an external PCIe interface. That's inherently insecure. It was bad enough that Firewire gave devices DMA access, but with PCIe it will probably be 10x worse.

But that's just a broken glass fallacy. It's nothing but waste.

They're more expensive than most of these boards, but it sounds like Soekris boards are more or less what you're looking for.

I'd say that them claiming any sort of copyright is indeed weird, because it's not really a creative work. IANAL at all, but it seems to me like random emails simply fall well outside the domain of copyright. For the most part, they aren't creative or artistic in any way. The only way they could possibly be copyrightable is if you made the argument that they are "literary" works, but that's a stretch.

Poor intake is the cause for most poor laptop cooling. On my Thinkpad W510 (which is quite poorly designed thermally, it's the same cooling at the T510 but it has to handle a quad core and workstation graphics), lifting up the keyboard a little bit gives a 10-15C drop in temperatures simply due to the fact that the fan will actually have intake.

Something tells me elementary school students already run into attention span issues, and making days longer certainly isn't going to help with that.

What I don't get is why they aren't addressing the real problem.

Normally, it can be good to have students of varying skill levels*, because the better students can help the lesser students. However, it seems that we often see stories about "Cheating in CS classes" here, which effectively discourages people from collaborating at all because it gets seen as cheating. Ideally, helping others should be rewarded, so that people who know the content would get their easy credit, while those who want to learn it would be able to do so more easily.

*Some education programs take this way too far. I've heard horror stories about how a high school geometry class had actual geometry students lumped in with special education students. Obviously that doesn't work but some overpaid idiot thought it would be a good idea anyway.