1,443 karma · joined March 23, 2009
[ my public key: https://keybase.io/kgrinberg; my proof: https://keybase.io/kgrinberg/sigs/WEI_d3HGjUIUBsOtsfBAyLlxPP2GxVWvgMAg5oxxwJE ]
Obviously you make a decision every time you give another entity any information, and storing data with Google/Facebook/et. al. carries risks as well.
But, as others have pointed out, the concentration and nature of this information matters too.
I'm reminded of the Veronica Mars episode about the "Purity Test" -
Step 1: get everyone's socially valuable data Step 2: Hilarity/blackmail ensues
Just to be clear - I'm not accusing the site of doing anything nefarious, just pointing out the difficulty of establishing trust.
I suspect there's a lot of money to be made in figuring out how to bridge that trust gap (of course, that's a little bit like the various authentication or email trusted-sender schemes that float about - IF ONLY we could will some universally trusted and known-good structures into existence, certain problems would be that much easier to solve...)
Consider, for instance, that for different groups, "PG" most likely means a company (Procter & Gamble), a movie rating, or some other industry-specific thing.
Obviously this isn't a particularly novel realization; perhaps it's even an "outsight" [1]. But it did strike me as interesting just how primed I was to expect a particular association for the relatively common name "Joel" - and reminded me once more of the importance of stepping out of the proverbial cocoon once in a while.
As to the merits of "form" classes and post handlers, I think the complexity is only there if you need it, if you're doing something more interesting (read: complex) than CRUD-type inserts/updates and basic validation (if that's all you're doing, ModelForms are pretty close to magical and light on syntax).
For instance, many travelers to Israel go for at least tangentially religious reasons; it's common practice for El Al security to ask those folks some basic questions about the faith (logic: if the guy claiming to be making a religious trip doesn't know something any observant pre-teen would, that guy deserves a closer look).
That doesn't mean that everyone traveling to Israel is Jewish (clearly not the case!), or that it's really all about profiling. But having a more homogeneous and constrained baseline (in addition to a smaller country) makes the "let's ask you some questions and see if your story makes sense" approach considerably more scalable.
Why does country size matter? A guy says "I'm visiting friends... where? Comstock - it's near Sargent." Without Google Maps at the ready, it would be hard to know if that's a plausible travel plan; whereas in a country that's only 8,000 sq. mi., with fewer people than NYC - a security agent is more likely to know shenanigans when s/he hears it.
It's sort of like the party game Werewolf (a.k.a. Mafia, etc.) - when played in person, it typically revolves around gestures, eye contact, and other physical factors; when played on a forum/over email, it's more about the psychology and strategy of who said what.
Or (and this is admittedly an area that I have less first-hand experience in), poker - past a certain point, the really good poker players tend not to give the game away based on physical tells, though the psychology and emotion is still a big contributor to what happens.
PEP 3003 puts the brakes (temporarily!) on one point of evolution - giving the rest time to catch up. Certainly, it hardly spells the death of the language.
Actually a pretty reasonable write-up of open source advantages.
http://blogs.msdn.com/oldnewthing/archive/2003/08/22/54679.a...
Unlike, say, Standard Oil, antitrust laws don't apply here (the problem isn't antitrust, it's systemic risk). So either you need some new laws, or just have to ask very nicely (though as a matter of future policy, regulators can - and should - refuse to approve these financial mega-firm mergers).
I suppose the government could have played a game of chicken (as it did with GM): "do what we tell you (break yourself up into smaller pieces) or we'll let you go under"... but if we've demonstrated that we weren't willing to countenance these institutions going under at that particular point in time, I'm not sure how credible the threat would be.
(Just to be clear, I'm not accusing you of making that argument, but the conflation of private and public debt is along those same lines).