What even totally zero-web-of-trust PGP gives you over "here's your
What even totally zero-web-of-trust PGP gives you over "here's your
Interesting coming from a company that will sell you a 3y9m old machine today (http://www.apple.com/shop/buy-mac/macbook-pro?product=MD101LL/A&step=config#). Reports are that they still sell rather a lot of them, because they're upgradable, repairable, and work just fine.
As for me, my 2010 MBP literally came out of a garbage skip. Found it with a bulging/burst lithium battery (far from an Apple-only issue). $50 worth of eBay grey market battery later, and I have a pretty solid machine for XCode and Mac testing. If it weren't for that, I just wouldn't test or dev anything for Macs. Couldn't afford to.
The Samsung Pay magnetic induction method is patented tech, from a company called LoopPay which Samsung bought.
AFAIK, they aren't licensing it to any other manufacturers at this time, because it gives them a clear distinctive advantage in an otherwise pretty-commodity marketplace.
NFC is also hugely more secure.
Newer Yubikeys (Yubikey 4) allows up to 4k RSA keys, as well as some elliptic-curve keys. Mind you, smartcard-based 2048-bit RSA encryption is wildly better encryption than 99.9% of the world. Especially if you're not really thwarting the NSA, 2k is FINE.
But in general you're absolutely right. Carrying around an easily copied keyfile is really no spectacular increase in security. Smartcards (where the decryption step happens on a completely separate micro-micro-processor, right there inside the same physical chip as the memory) is by far the best answer we've got.
As for the question on safekeeping? The extreme paranoia method is to generate your keys on a totally airgapped (no network) old laptop and save the backup private keys in an encrypted volume. Then copy the private keys onto a smartcard/Yubikey for daily use (most smartcards allow a one-way push of a private key from PC->smartcard, but no retrieval of the private key).
What does a decently-spec'd MBP or non-Apple equivalent and a mid-grade commercial IDE cost these days? $2k at the most?
Even if you are hiring 18-year-olds in rural South Dakota, you are looking at $50k a year ($35k salary + other direct costs), the $2k is *nothing*. 16GB laptops, SSDs, giant screens, and huge backup arrays are close to nothing.
Don't be that cheap guy. Don't work for that cheap guy. If you are you own boss and are the one cheaping out on your own self... look in a mirror, do a Stuart Smalley Daily Affirmation and step away from cheapness-first.
Where does this penny-pinching come from in IT? I think I know where it comes from, because I was there. Let's take a random year like 1989. A new 'fancy' machine, like, say a Mac IIci with a color monitor was over $10,000 ($20k in 2016 dollars). No joke. Half a year's salary. If you were able to do much of the same stuff on a cobbled-together PC/AT clone for a third of the price, you were ahead. Great. I grew up poor, with a bunch of nerdy poor friends, and we were scrambling to put together thrown-out old Zenith 8088s. Great.
It's not 1989. Especially when you compare, say, a two-year-old Thinkpad for under $500 to 'making things work, mostly' on a slightly cheaper Chromebook (I'd want the former, no question)? It's Just. Not. Worth. It.
You want your developer to have the oomph to play with VMs and Docker and whatever cool crap comes out tomorrow.
As for everything else? With all my tooling scripted up and in version control, I can go from an Ubuntu iso to basically fully operational in about 90 minutes. Might suck a lot more with Windows fiddle-twiddling (especially if you're not big enough for images and domain-centric centralized management). Syncing, teamwork, and deployment are already covered... and it's not the IDE's bailiwick.
Right with you there. Look, I've been a Linux user, one way or the other, for even a little longer than that (Slashdot ID checks out). I've been whatever-coexisting with Windows for the last decade or so. The period where sound and wifi were sucking on Linux (and IE ruled the web) coincided with me having enough income to buy new-out-of-box laptops. So, grew to live in a Windows desktop, Linux server peace. Actually didn't hate Win8/8.1 for my own needs (though I agree it was a UI disaster for non-power-users).
Between the Win10 spycrap and the nag screens, though, I finally said 'fark it'. I'm back to 100% desktop Linux, 100% of the time, for the first time in over a decade. It's really, really refreshing.
If it's legally impossible to request a review of them, why bother with creating and storing the paper tapes in the first place?
Which leads, I guess, to the next question. If it's legally impossible to review an election, why bother holding them in the first place?
Early 1995, still in high school. I was in a small town in Kansas. Absolutely disconnected from the pre-web internet. No BBSes or anything that wouldn't be a long-distance call. And my parents were fairly poor (okay... lower-middle income but horrible with money), so no long distance.
But geeky. My dad bought into the TI-99 after TI pulled out of the home computer industry because he could buy a computer for $50. There was a whole community of people who did fairly amazing things with 15-year-old hardware well into the mid-1990s (heck, there's still a few around today, like old Atari/Amiga/Apple ][, etc groups). One day, along with the shareware TI 5-1/4" floppies that we were mailing around with other users, there was a Slackware CD. I had recently scrounged together a 486 that was capable of running it. And Bob's your uncle.
Really addicted to mine. I have my private SSH key on there (via GPG/PGP), so that's never on my working machines. Use the standard OTP on several personally-run sites. Use U2F security for Google apps. Use the TOTP (a.k.a. Google Authenticator/Authy) app. Use the challenge-response mode as a second factor on my KeePass database. Amazing gadget.
The question regarding the teardown is... "so"? Even with full pin access to the A7005 chip, you *STILL* wouldn't have access to my GPG/SSH private key or my TOTP generators within it. That's the point of a secure element. You'd have to dissolve the casing of the A7005 chip and have a decent microscope lab to get those bits of data out of the chip. You would be able to use my U2F/OTP/TOTP-generated-code functionality. But, you could do that just by stealing my Neo and plugging it into a USB slot without any acetone bath involved.
As said above, SSA doesn't have any sort of biometric verification of "who you are".
And, as said above, your SSN shouldn't be used as an identifier. If we need a common citizen ID number, fine, but it shouldn't be anything but identifying (i.e., effectively public knowledge).
It's the gorram 21st century. We've had public-key encryption figured out for over 30 blessed years now. Most people in the first world are carrying around several crypto smartcard devices already (EMV compatible credit cards and other smartcard tech).
Much of the world now has ID cards with cryptographic chips in them. When you open a line of credit, you prove, through RSA/elliptic-curve signatures that you are YOU via your ID chip. If you lose your ID, it gets put on the centralized revoke list, the issuing agency goes through whatever in-person process to verify you are you, and gives you a new ID. This can extend to online purchasing, online voting, etc, etc.
But, we're so freaked out about government black helicopters that we just accept the whole fraud thing as inevitable.
Is it, though? It was infinitely easier to carefully (okay, obsessively) portion out the 1700 kCal per day I could eat and maintain just-under-obese status when I was single and nearly a hermit. Married (to a gal with better metabolism than me), there's simply endless, "hey, I made cookies" or "hey, I'm just springing on you that we're going out with friends for fish-n-chips tonight" temptations.
The flip side of the question is "Why are skinny people not fat?".
It's a more interesting question than you may think. One bit of semi-famous research is the 1970s Vermont 'prisoner overfeeding study' (http://www.youtube.com/watch?v=P5Rv8JnFgw4). Like bits of Nazi science, this is probably irreproducible, as it'd *never* get past a human subject review committee today.
A number of lifetime-normal-weight prisoners were fed substantially over their basal metabolic needs for an extended period. Their input was rigorously controlled (being prisoners), and their exercise regimen was pretty easy to monitor and control. Most of them gained weight, but almost none of them nearly as much as the standard "3500 kCal is a pound of fat" Standard Model would predict. Several plateaued on weight gain, and a few lucky (?) prisoners were *never* able gain 10% of their body weight when eating nearly 10,000 Calories a day. Simply couldn't do it.
A lot of people are overeating in the western culture. A lot more that, by the numbers, should be in the 300-pound range. And while there are no shortage of very-very-fat people, they're not nearly as common as they should be if you study individual diet patterns. This is part of the problem. People look at their skinny friends' diets, and some of those skinny friends are like the luckier Vermont prisoners.
The basic principle is sound. 6 days income for 15mph over is pretty stiff, but then again, a lot of US speeding tickets now are in the $400+ range (after court fees, etc) which is 6 days' income for a lot of people.
"15 mph/25 km/h over" is kinda a poor starting point. 55km/h in a 30km/h zone (one that really needs to be a 30km/h zone... like a dense urban center with playgrounds and schools)... to me that's pretty deserving of punishment. 125km/h on a rural road posted at 100km/h in clear weather? I'm not sure that even merits a warning. I'd put the penalties at 30%/40%/50%/60%/70% over the posted limit rather than a fixed speed-delta.
Value judgement time, but for my money, nobody's out there brute-forcing RSA keys even at 1024-bit except, maybe, the NSA. If you weigh "everyone but the NSA" security as a bigger day-to-day concern, side-channel issues (keylogging, shared memory, copied private key files, implementation flaws, etc) are a lot more pressing realities than the almost-theoretical added security of 4kb+ RSA keys or going ECC.
One bit of paranoia the author might add is moving your private key completely off of your desktop into a smartcard that does the RSA or ECDSA step and, being a far more limited microprocessor, should be more securable than processes running on a general-purpose networked computer and multitasking OS.
I believe there are ways to do ssh with PKCS-based smartcards, but the method used around here is based on PGP/GPG keys and either the "OpenPGP Smartcard" (ISO smartcard form factor, requires a smartcard reader) or the YubiKey Neo (USB pen-drive form factor). You create a key pair (possibly using the smartcard CPU itself). You use gpg-agent with OpenSSH (or PuTTY) support instead of ssh-agent/pageant. The private key never leaves the device (the little bit of flash memory in the chip) and is designed to be unrecoverable. The RSA authentication step happens in the microprocessor on the card. The card has a PIN and is designed to lock after a couple missed PINs.
http://www.bradfordembedded.co... for a starting point.
A man is known by the company he organizes. -- Ambrose Bierce