1) to control access to data the user cares about
2) to externalize the costs of controlling access to data the company cares about onto the user

123456, password, etc. are perfectly valid and rational user responses to the latter situation.

Because, of course, it is so much better to sell your users to some social network and let them control how you run your site or business?

Webmasters do live in and manage their own universes, to the extent that they want to. What next, you're going to complain I have a door on my house or on my bathroom? Go away, you're creepy.

and a silly suggestion.

How many bits of entropy are you actually producing? If you don't know, go to the back of the class.

n/t

"You must pay us upfront before you even start if you even think you might commercialize your software" is the acme of desperate opportunism (aka shark-jumping) and of an inflated sense of self-importance, and probably entirely unenforceable as a matter of copyright.

A BSD-licensed toolkit of usable quality would, indeed, correct Qt's attitude problem.

As to quality, have you ever built and ran gimp-1 prereleases? "Interesting" times....

You're implying that the political act of buying a certificate makes one's bloviations on any particular subject more credible. Which, like all elite outgroup favoritism, is the position of the supplicant and the traitor.

The marginal value of a degree has nothing to do with education. It's the debt and the acculturation to bourgeois Whig values that employers consider desirable.

makes a fine covert channel to get data to or from a compromised router, and NSA has shown interest in mass-pwning routers.

Well, good. Religion has no place outside of religion, and certainly not in math.

They have lots and lots of things for every occasion. And remembers, keys do not only encrypt, but authenticate. How does it feel having a copy of one's car keys left with the NSA, especially post-Hastings?

Apt Richelieu quote. They are collecting data on every person, but how much is a function of the person (or the data). Consistent sympathizers with incumbent power are uninteresting and not really worth the bandwidth. Sympathizers with any power other than the regime are interesting.

What information would they have gotten by holding a competition that they wouldn't have gotten by developing internally?

Competitive intelligence, duh. If you want to see what the state of the art in civilian crypto is, of course you want to look at as much as possible.

Either way, that's kind of a strange thing to say about an organization that claims to have completed an awesome new cryptanalytic capability in 2011, after which (according to the black budget leaks) CCP's Microelectronics program shrunk by a factor of six over the next two years... and that slide with that little red box...

Look, if you're reading this news site, and you didn't know that security can only be meaningfully assessed against a threat and that an accurate model of this threat is required in order to assess it and propose countermeasures, you're not going to bring much to this conversation except noob blather.

The observation you appear to be missing is that *both* adversaries in a crypto war have cost-benefit tradeoffs to make. Backdooring a crypto product is cheap, and very valuable. Expect it. The NSA black-bagging your hardware is expensive. If you've never been to an environmental protest, you're not worth this to them. The NSA pwning your router, apparently, is a capability they hope to automatically acquire this year -- expect that.

THIS. Americans relate to duplicity like fish relate to water: it's apparently the medium of action in the world we inhabit, yet we can't see it.

Might makes right, eh? I take comfort in knowing that the smarmy cosmpolitan bourgeois Randites will be among the first against the wall.